中文
Founder Stories Collection

Founder Stories of
America's Top 200 Companies

From NVIDIA to W.W. Grainger — Behind every great company is an extraordinary beginning

200Stories 200Founders 200Companies
▼ Start Reading

Contents

Click any chapter to jump

  1. 001 NVIDIA 32 years, $5 trillion. He did only one thing: bet on a future nobody believed in.
  2. 002 Google 26 years, $4.6 trillion. They did only one thing: refused to sell a search engine nobody wanted to buy.
  3. 003 Apple Fired from the company he founded for 11 years. He came back and did something even more ruthless: killed 70% of the product line.
  4. 004 Microsoft One magazine cover, two Harvard dropouts, and a contract that changed the world.
  5. 005 Amazon Nine straight years of losses, stock cratering from $113 to $6. He did something crazier: kept investing.
  6. 006 Broadcom He acquired 40 companies — none of them his own — and built a $2 trillion chip empire.
  7. 007 Tesla Christmas Eve 2008. He had $200,000 left to his name — and he burned every last dollar into Tesla.
  8. 008 Meta One night at age 19, he built a website in his dorm room. Twenty-one years later, 2.7 billion people use it.
  9. 009 Berkshire-Hathaway 60 years, $1 trillion. He did only one thing slower than everyone else: nothing.
  10. 010 Walmart He opened 1,000 stores, every single one in a small town. All the city experts said: "Nobody buys anything in places like that."
  11. 011 Eli-Lilly At 38, a veteran army medic put a label on a pill bottle — and rewrote an entire industry.
  12. 012 Micron Three engineers built memory chips in a basement. Japanese giants pushed them to the edge of bankruptcy — so they did one thing: no money, keep building.
  13. 013 JPMorgan-Chase Panic of 1907. Every bank was fleeing with cash. He shut his library door, did the opposite — summoned America's biggest bankers and ordered them to pool their money to save the system.
  14. 014 AMD Intel chased him for 50 years. Every time he was nearly dead — he got up and kept fighting.
  15. 015 Exxon-Mobil 1870, Cleveland, Ohio. A 31-year-old bookkeeper called everyone into his office and said: from today, we do one thing — drive down the cost of moving every barrel of oil.
  16. 016 Visa 1958. A bank mailed 60,000 credit cards to every resident of a small California town. Nobody applied. Nobody activated. Every cardholder became, overnight — a victim of fraud.
  17. 017 Intel Eight engineers quit together and founded Intel. The most ruthless among them later wrote a book called "Only the Paranoid Survive."
  18. 018 Johnson-Johnson 148 years ago, he opened a pharmacy. His secret weapon wasn't medicine — it was an unbreakable principle: patients first.
  19. 019 Oracle At 35, he had zero money. He took a paper IBM left by the roadside and built a database — then he chased down his customers one by one, shouting: if you don't pay, you're a thief.
  20. 020 Costco He slashed retail gross margins to 14% — then told every board member who wanted to raise prices: anyone who raises prices leaves.
  21. 021 Mastercard 1966. A few banks sat down and decided to build an "anti-Visa alliance." Today it's worth $445 billion.
  22. 022 Caterpillar 1904, a marsh in California. A steam tractor sank into the mud, wheels spinning in slurry. The farmer crouched nearby, cursing. The photographer behind him, Benjamin Holt, didn't help drag it out — he knelt and stared at the wheel. Only one question in his head: what if a wheel wasn't a wheel?
  23. 023 Cisco On the Stanford campus, a couple built a router in a lab — because different department buildings couldn't email each other. Their department chair said it had no commercial value. Fifteen years later, Cisco was the most valuable company on Earth.
  24. 024 Netflix 1997. He rented Apollo 13, returned it six weeks late. $40 late fee. He walked out of the video store with one thought: I'm going to kill this industry.
  25. 025 Chevron 1911. The U.S. government carved Standard Oil into 34 pieces. One piece was called Standard Oil of California — today, it's Chevron, worth $370 billion.
  26. 026 AbbVie Some called him the world's most profitable "patent thief." But every company he "stole" from ended up making even more money.
  27. 027 Bank-of-America 1906 San Francisco earthquake. The entire city burning. Every bank frantically hauling gold bars out of town. Only one man — a small Italian immigrant — ran the opposite direction into the ruins. He stuffed $80,000 in gold bars into a vegetable wagon, covered them with oranges, and drove through burning streets back to his little shop. The next morning, he set up a table on the wharf.
  28. 028 Lam-Research 1980, Silicon Valley. An engineer who'd spent over a decade etching chips at Fairchild and Intel — David Lam — sat at his kitchen table and told his wife: "I'm quitting. I'm going to build something everyone says nobody needs — a machine that does only one thing: etch." At the time, the entire semiconductor industry was selling all-in-one machines. He bet everything on one single process.
  29. 029 UnitedHealth 1974, Minnesota. A group of doctors argued in a conference room in suburban Minneapolis. The core debate was a sentence that sounded heretical at the time: "Can we let doctors — not insurance companies — decide what treatment is 'medically necessary'?" The man who said it was Richard Burke. He wasn't a doctor. He was a health policy economist — and he built the company that would become UnitedHealth.
  30. 030 Coca-Cola 1886, Atlanta. A pharmacist boiled a pot of brown syrup in his basement. He sold it as medicine — for headaches, fatigue, neurasthenia. First year, he sold just 25 gallons. But then someone turned that pot of syrup into the most famous taste on Earth.
  31. 031 Procter-&-Gamble 1837, Cincinnati. An English immigrant candle-maker and an Irish immigrant soap-maker shared a delivery driveway. They married a pair of sisters — then their father-in-law called them to his table and said: "You're family now. Merge." P&G was born. Today it's worth $337 billion, and its products sit in virtually every home on Earth.
  32. 032 Applied-Materials 1967, Santa Clara, California. A chemical engineer named Michael McNeilly — out of TI's semiconductor materials division — set up a small lab in his garage and began using chemical reactions to "grow" an ultra-uniform insulating film on silicon wafer surfaces. He named the company Applied Materials. Today it is the world's largest semiconductor equipment manufacturer.
  33. 033 Palantir After 9/11, a Stanford philosophy PhD and a PayPal mafia member sat down and made a decision: we're not building an AI assistant. We're building a counterterrorism platform to "find terrorists." The U.S. government became their first — and for over a decade, their only real — customer.
  34. 034 Home-Depot 1978. Two 49-year-old men got fired from the company they founded. They sat in an Atlanta coffee shop. One of them said: what if we opened a hardware store — better than the one that just kicked us out?
  35. 035 General-Electric Thomas Edison founded GE. But his company later did something he could never have imagined — 100 years after his death, GE split into three companies. And the filament inside Edison's original incandescent bulb is still burning today.
  36. 036 General-Electric 1892, Schenectady, New York. Two inventors held a tangled mess of patent lawsuits — you sue me for infringing the light bulb patent, I sue you for stealing AC transmission. Behind one stood J.P. Morgan, behind the other the Vanderbilt family. Morgan summoned both sides to his study and said: "From this moment, you no longer exist — only one company exists."
  37. 037 Morgan-Stanley September 16, 1935, 2 Wall Street. A napkin hit the table. The man who threw it was Henry Sturgis Morgan — grandson of J.P. Morgan. He stood up and announced to the Morgan partners: he and three others were leaving Morgan Bank to start a new firm — one that would never lend money to companies, only advise them.
  38. 038 Philip-Morris 1847, Bond Street, London. A shopkeeper named Philip Morris opened a small retail store. He didn't sell "cigarettes" — they hadn't been invented yet. He sold hand-rolled pipe tobacco and imported Turkish blend. His shop door bore a handwritten wooden sign: "Philip Morris — Tobacconist & Barber."
  39. 039 GE-Vernova April 2, 2024. After 132 years, General Electric split from one company into three. One of them is GE Vernova — inheriting all of GE's energy businesses: gas turbines, wind turbines, grid equipment, hydro, and nuclear. On its first day trading independently on the NYSE, it was already one of the largest energy equipment companies on the planet.
  40. 040 Goldman-Sachs 1869. A German Jewish immigrant hung a brass plaque in a basement on Pine Street, New York: Marcus Goldman — Commercial Paper Dealer. He was 48, with only a leather satchel and an address book. Every morning he took a stagecoach to jewelers and leather merchants upstate, buying their promissory notes with cash, then selling them to downtown banks that same afternoon.
  41. 041 Texas-Instruments Summer 1958, Dallas. All Texas Instruments employees were on mandatory two-week vacation. Only one new hire — Jack Kilby — had no vacation, having started just three months earlier. Alone in the empty lab, he did one thing: etched transistors, resistors, and capacitors onto a single piece of germanium crystal. The world's first integrated circuit was born in that silence.
  42. 042 RTX---雷神技术 1922, Cambridge. Three MIT engineers — Laurence Marshall, Vannevar Bush, and Charles G. Smith — founded Raytheon in an old warehouse. Its first product wasn't a missile — it was a home refrigerator compressor. But it didn't cool. They turned the failed compressor into an electronic component that replaced vacuum tubes. Today Raytheon is a $150+ billion defense technology giant.
  43. 043 KLA---科磊 1997. An MIT-educated engineer, Ken Levy, became CEO of the merged KLA-Tencor. He ran an "invisible company" — its machines don't make any chips; they inspect chip manufacturing for defects invisible to the human eye. If TSMC and Intel's lithography machines are the paintbrushes, KLA is the microscope that checks every single stroke.
  44. 044 Wells-Fargo 1852. Two Eastern merchants stepped off a boat at the San Francisco wharf. The entire city was gold-rushing — everyone was in the mountains digging. They didn't go to the mountains. They put a table on the dock with a sign: your gold dust stays here — and you'll find it here tomorrow.
  45. 045 QUALCOMM 1985, San Diego. Seven engineers started a company in a garage. Their first product was a vehicle-mounted satellite positioning terminal — big, expensive, no customers. Then they did the reverse: stopped selling terminals. They sold "the math inside the head." They turned a military spread-spectrum technology called CDMA into the soul of every cellphone on Earth.
  46. 046 American-Express 1850. Three stagecoaches departed Buffalo, New York. They weren't carrying letters — they were carrying money. The safes onboard were packed with wax-sealed cash and gold. A hand-painted wooden sign hung from the lead coach: "American Express." Today it's worth $213 billion.
  47. 047 Citigroup 1986. A Brooklyn-born stockbroker sat in his Manhattan office and hung a crayon-drawn org chart on his wall: at the top, "Us" — a consumer loan company too small to matter. Below, a string of names he planned to devour: American Express, Aetna, Citicorp. He would eventually take over Citicorp itself — and build Citigroup.
  48. 048 T-Mobile-US 1999. Deutsche Telekom's CEO made a decision every board expert shook their head at — spending $35 billion on a tiny U.S. carrier called VoiceStream. It had fewer than 2 million customers and coverage full of holes like Swiss cheese. They called it "sucker's tuition" for entering the U.S. market. Today T-Mobile US is one of America's Big Three wireless carriers.
  49. 049 SanDisk---闪迪 1988. A 65-year-old Israeli immigrant founded a company called SunDisk in California. His hair was all white. His accent was thick. He'd already failed twice before founding this one. Investors said: "You're starting a company at retirement age?" He smiled and nodded. Today SanDisk is worth $208.7 billion.
  50. 050 IBM 1914. A 40-year-old tabulating-machine salesman walked into a company called CTR. Everyone was doing their own thing — tabulators, time clocks, scales — three piles of machines that didn't talk to each other. He did something nobody did: he borrowed an enormous sum, signed a contract deeply unfavorable to himself, and staked his whole life on this company. His name was Thomas J. Watson.
  51. 051 PepsiCo 1893, New Bern, North Carolina. A pharmacist named Caleb Bradham mixed a brown sugar drink in his basement. He called it "Brad's Drink" — for indigestion. His neighbors said the burps felt great. He renamed it Pepsi-Cola. Today PepsiCo is worth $207.4 billion.
  52. 052 Blackstone 1985. Two Lehman Brothers partners each put up $200,000 and rented a small office in Manhattan. At their first morning meeting, one looked at the other and said: we will never do a hostile takeover. The world already has enough people stealing from each other. What we'll do is — buy companies, make them better, then sell them. Today, Blackstone is the world's largest alternative asset manager.
  53. 053 Verizon 2000. Bell Atlantic acquired GTE for $64.7 billion — then did something that baffled Wall Street analysts. They threw both names in the trash and coined a new word: Verizon — fusing "Veritas" (Latin for truth) and "Horizon." Today Verizon is one of America's telecom giants.
  54. 054 Analog-Devices 1965. Two MIT graduates — Ray Stata and Matthew Lorber — founded a company in an old Cambridge garage. They didn't make digital chips, DRAM, or microprocessors. They made something the all-digital age considered "old-fashioned" — analog chips. Converting real-world temperature, pressure, and sound waves into precise digital signals. Today ADI is a pillar of the semiconductor world.
  55. 055 McDonalds 1954. A 52-year-old milkshake-machine salesman drove three days from Chicago to San Bernardino, California. He'd received a strange order — a hamburger stand bought 8 milkshake machines. Eight machines meant 40 milkshakes at once. He had to see it with his own eyes. He arrived at noon and saw an assembly line he'd never seen before. His name was Ray Kroc.
  56. 056 NextEra-Energy---新纪元能源 1925, Florida — still a swamp and orange groves. A tiny utility called FPL (Florida Power & Light) launched in Miami, its earliest generators rebuilt from diesel engines salvaged from shipwrecks. A hundred years later, its parent NextEra Energy is worth $196.9 billion and is the world's largest producer of solar and wind energy.
  57. 057 Disney 1928. He lost his first successful cartoon character — Oswald the Lucky Rabbit. The market didn't beat him. His distributor had buried a clause in the contract: Oswald's copyright didn't belong to him. He sat on the train from New York back to Los Angeles and drew a mouse on a napkin. Today that mouse is worth $182 billion.
  58. 058 Boeing 1916, Seattle. A Yale-educated timber merchant sat in his first seaplane — a biplane stitched together from spruce wood, linen, and piano wire, with a wooden propeller and a repurposed boat engine. He test-piloted it himself. The moment it touched down on Lake Union, he said to his partner: we're going to build the best airplanes in the world.
  59. 059 Amgen---安进 1980. A venture capitalist named Bill Bowes and a handful of scientists founded AMGen (Applied Molecular Genetics) in Thousand Oaks, Southern California. They had no drug molecule of their own, no factory — only a theory: recombinant DNA could "print" human protein drugs from E. coli bacteria.
  60. 060 ATT March 10, 1876, Boston. Alexander Graham Bell sat in his rented lab and spilled sulfuric acid on his leg. He shouted into a strange metal apparatus — not to talk to anyone. He was calling his assistant: "Mr. Watson — come here — I need you!" Watson wasn't in the room. He was in the basement, and he heard every word through the wire. That was the first phone call on Earth.
  61. 061 Arista-Networks 2004. A former Cisco engineer, Andy Bechtolsheim, founded Arista in Palo Alto. He'd discovered a fatal crack Cisco overlooked while building data centers for Google and Microsoft: existing switch operating systems were mountains of patches stacked on 1980s code — and cloud data centers needed a "programmable, flat, software-driven" network.
  62. 062 Palo-Alto-Networks 2005. An Israeli-American engineer named Nir Zuk sketched a diagram in Sequoia Capital's Sand Hill Road office — he said a firewall isn't about "ports." Real security must know whether each data packet carries Facebook, Gmail, or malware. His former colleagues at Check Point said he was wasting time. Today Palo Alto Networks has a market cap in the hundreds of billions.
  63. 063 Thermo-Fisher---赛默飞 1903, Chicago. A scientific instrument salesman named C.G. Fisher opened a small supply house in a downtown attic. He sold just one thing: a precision analytical balance that let university chemistry departments weigh to one ten-thousandth of a gram. Over a century later, Fisher Scientific merged with Thermo Electron — becoming Thermo Fisher Scientific, the world's largest life sciences tools company.
  64. 064 Western-Digital 1970, Santa Ana, California. An engineer named Alvin B. Phillips founded Western Digital — initially making semiconductor test equipment. Then it pivoted to storage controllers, then hard drives. In 2016, it acquired SanDisk for $19 billion, becoming a dual-engine HDD + NAND flash powerhouse. Today WDC is worth $168.4 billion.
  65. 065 Gilead-Sciences---吉利德科学 1987. A doctor named Michael Riordan founded a small biotech company in Foster City in the San Francisco Bay Area. He wrote three words on a whiteboard: "Cure. Don't control." — he was targeting viral hepatitis and HIV. Later, Gilead's Sovaldi would become a cure for hepatitis C, making over $10 billion in a single year and changing the course of an entire disease.
  66. 066 BlackRock---贝莱德 1988. Larry Fink founded Blackstone Financial Management in a small office in Manhattan. His first big loss was massive — an interest-rate bet at First Boston had cost a client $100 million and nearly destroyed his career. Then he decided: no more betting on direction. He would build a machine that "quantifies all risk with math." Today BlackRock manages over $10 trillion.
  67. 067 Corning---康宁 1851. A businessman named Amory Houghton bought a bankrupt glass factory in Corning, New York. Using the local quartz sand — some of the purest natural silica on Earth — he began making red glass lenses for railroad signal lamps. That glass factory became Earth's greatest specialty glass innovation engine. Today Corning is worth $167.3 billion.
  68. 068 TJX-Companies 1976. Bernard Cammarata opened the first T.J. Maxx in suburban Boston. He sold what other department stores left behind — last season's clothes, odd sizes, manufacturers' excess inventory. He bought low and sold the same brands at 70% off. Department stores called him "a flea market for cheap branded knockoffs." Today TJX is the world's largest off-price retailer.
  69. 069 AppLovin 2012, Palo Alto. Adam Foroughi founded AppLovin, a mobile ad platform. Its logic was the reverse of everyone else — not helping brands run ads, but helping app developers cross-promote users to each other. A small game ad promoting another small game — optimized by machine learning. Then he built his own game studios, cycled his own users through his own algorithms.
  70. 070 Deere---迪尔 1837, the Illinois prairie. A blacksmith named John Deere moved from Vermont. Pioneer farmers used cast-iron plows designed for Eastern sandy soil — but the Midwest was sticky black chernozem. Iron plows sank in like they'd hit asphalt; farmers stopped every few steps to scrape mud with a stick. He took a discarded sawmill blade, cut out a mirror-polished steel plow, and reshaped farming on the American frontier.
  71. 071 Uber Winter 2008, Paris. Two tech entrepreneurs stood beneath the Eiffel Tower for 45 minutes unable to hail a single taxi. Snow was falling. One looked at the other: "Wouldn't it be cool if you could press a button and a car would come?" The other said: that's a limo service, ridiculously expensive. The first replied: "What if it was for everyone?" Today Uber is a global mobility giant.
  72. 072 Charles-Schwab---嘉信理财 May 1, 1975 — the very day "May Day" deregulation took effect — a 34-year-old stockbroker named Chuck Schwab registered a discount brokerage called Charles Schwab in San Francisco. His office had one folding chair, one steel desk, one phone. No analyst reports, no investment advice. Just one thing: buy and sell stocks for people at ultra-low commissions.
  73. 073 Uber A cold winter night in 2008, beneath the Eiffel Tower in Paris. Garrett Camp and Travis Kalanick stepped out of a tech conference party — rain falling. They stood in the street trying to hail cabs for 30 minutes. Camp turned to Kalanick: "My dream right now — pressing a button on my phone and a car shows up."
  74. 074 Welltower 1970, Toledo, Ohio. A real estate investor named Fred Carr made a decision in healthcare REITs that baffled every property developer at the time — he would only buy hospital buildings and senior care facilities, then lease them to healthcare operators. His entire life, he did one type of property: "buildings where sick people go to heal." Today Welltower is worth $153+ billion.
  75. 075 Intuitive-Surgical 1995, a Stanford Research Institute lab. A group of scientists was studying "how to let surgeons operate without putting their own hands directly inside the patient's body." Then Intuitive Surgical's founder Dr. Frederic Moll turned it into a robot called da Vinci — translating a surgeon's microscopic, tremor-free finger movements into motion inside the patient's body through four robotic arms.
  76. 076 Dell 1984, a UT Austin dorm room. A 19-year-old named Michael Dell began selling custom-assembled, upgraded IBM PC compatibles — cheaper than stores, built to order. His roommate hid his box of parts in the shower stall. Today Dell Technologies is worth $151.2 billion. His core principle — "direct sales, build-to-order" — is now a manufacturing religion.
  77. 077 Intuitive-Surgical 1986, a basement lab at Stanford Research Institute. A microelectronics engineer named Dr. Phil Green was performing microsurgery experiments under an electron microscope with a needle nearly invisible to the naked eye. His problem wasn't "can I cut tissue" — he certainly could. His problem was: "The human hand trembles. Under the microscope, mine trembles 0.3 millimeters." He began designing a machine that could subtract that tremor.
  78. 078 Southern-Copper---南方铜业 1952, in the Arequipa valley of southern Peru, America's Asarco Copper and the Peruvian government jointly opened the Cuajone and Toquepala mega open-pit copper mines. What began as a conventional multinational-plus-state contract evolved through mergers and multi-country Latin American operations into one of the world's largest copper companies.
  79. 079 Pfizer 1849, Brooklyn. Two German cousins — Charles Pfizer and Charles Erhart — opened a chemical workshop on the second floor of a red-brick building. Their first product was a white powder that tasted like sugar but killed intestinal parasites. They mixed it with toffee — creating the world's first "medicine that tasted good." Today Pfizer is one of the world's largest pharmaceutical companies.
  80. 080 Abbott-Laboratories---雅培 1888, Chicago. Dr. Wallace C. Abbott began producing something called "Abbott Granules" in his apartment kitchen — placing alkaloids into extremely precise micro-dose granules so doctors on house calls could give patients exact doses without a scale. He was a practicing physician, not a pharmacist — but he saw patients suffer daily from dosage errors and decided to fix it.
  81. 081 Salesforce 1999. A former Oracle vice president founded a company called Salesforce in his apartment. His first investor, Larry Ellison, gave him $2 million. But he did something even Ellison found outrageous — during the Super Bowl he rented a plane to fly over Silicon Valley trailing a banner, and at every company party he sent fake protesters with signs: "SOFTWARE IS DEAD."
  82. 082 ConocoPhillips---康菲石油 1875, Ogden, Utah. A small company called the Continental Oil and Transportation Company began hauling kerosene between Western states — using a fleet of horse-drawn wagons and wooden barrels. Today its descendant ConocoPhillips is worth $141.6 billion. Conoco is the quintessential independent oil exploration and production company — no gas stations, no refineries, no chemicals. Just one thing: finding oil and gas beneath the Earth's surface and bringing it up.
  83. 083 Marvell-Technology 1995, Berkeley. Sehat Sutardja, his wife Weili Dai, and his brother Pantas Sutardja founded Marvell Technology at their kitchen table. Their first self-designed chip — a fully digital hard-disk read channel — consumed one-tenth the power of the industry standard. Today Marvell is worth $141.3 billion, having expanded from storage controllers into data center networking and AI infrastructure silicon.
  84. 084 Interactive-Brokers---盈透证券 1977. Budapest-born Thomas Peterffy sat in the corner of the New York Commodity Exchange. He was the first person on Wall Street to replace traders' shouted prices with computer models. He plugged an Apple II into a trading desk — while others were still using chalk and hand signals, he was computing option fair values in code. Today his company Interactive Brokers is worth tens of billions.
  85. 085 CrowdStrike 2011. George Kurtz and Dmitri Alperovitch founded a security company called CrowdStrike. They made a crazy bet — all antivirus software that relies on "virus signatures" to identify known malware was dead in the face of cloud-native and nation-state APT attacks. They didn't sell antivirus. They put a lightweight sensor on every protected machine, streamed all data to the cloud, and hunted threats in real time.
  86. 086 Honeywell---霍尼韦尔 1885, Wabash, Indiana. Albert Butz invented a thermostat — a device that automatically regulated the temperature difference so "cold air could rise from the basement to the living room." Then Mark Honeywell improved the patent and founded Honeywell Heating Company. Today Honeywell is worth $137.5 billion. Its control sensors live in every smart building, jet engine, and industrial plant.
  87. 087 Vertiv-Holdings 2016, Vertiv spun off from Emerson Electric and went public. But its true origin is the Liebert Corporation of the 1940s — founded by Ralph Liebert, who invented the first precision air conditioner for computer rooms. In 1965, he built the first "environmental control unit" for mainframe computers — cooling not for people, but for IBM mainframes. Today Vertiv powers and cools the world's data centers.
  88. 088 Prologis---安博 1984, San Francisco. Hamid Moghadam began investing in logistics warehouses at AM Property. Later he merged with Security Capital to form Prologis. Today Prologis is worth $134 billion — it is the world's largest owner of "the big metal boxes your Amazon packages sit in before reaching your doorstep." Every "next-day delivery" you click runs through a Prologis shed.
  89. 089 S&P-Global 1860. Henry Varnum Poor published the first "History of the Railroads and Canals of the United States." He was an eccentric who standardized and publicly released the financial data of American railroad companies — letting investors, for the first time, compare the financial health of different railroads. His book inspired the logic behind all financial information services. Today S&P Global is worth tens of billions.
  90. 090 Lowe's---劳氏 1921, North Wilkesboro, North Carolina. A small-town hardware store owner named Lucius Lowe turned his little shop into a one-stop building supply and home improvement store. A hundred years later, Lowe's Companies is worth $126.3 billion — the second-largest home improvement retailer in America, just behind Home Depot. The difference between them is subtle — and it all hangs on the distinction between professional contractors and DIY families.
  91. 091 Newmont---纽蒙特 1921. Colonel William Boyce Thompson founded Newmont Mining Corporation in New York and immediately bought gold mines in California and Nevada. A hundred years later, Newmont is the world's largest gold mining company — worth $125.1 billion. Its assets were geologically formed 200 million years ago in hydrothermal rock layers — and Newmont dug them out of the Earth's crust and melted them into bars.
  92. 092 Booking-Holdings 1997. Jay Walker registered Priceline.com in Stamford, Connecticut — a revolutionary name-your-own-price model. You didn't pick a flight — you said "New York to Miami, $100, any time" and airlines anonymously bid their leftover seats. It upended airline pricing. Later Priceline acquired Booking.com, renamed itself Booking Holdings, and became one of the world's largest online travel platforms.
  93. 093 Starbucks 1983. A Brooklyn-born kitchenware salesman flew to Milan for a housewares trade show. He walked into his first Italian coffee bar near the Duomo and stood at the counter watching for three minutes: a barista turned espresso and steamed milk into a brown whirlpool in under 60 seconds, then slid the cup to a man in a suit. The man downed it in one gulp. The salesman thought: we're not in the coffee business — we're in the "third place between home and work" business. His name was Howard Schultz.
  94. 094 Lockheed-Martin 1926, Los Angeles. A pair of brothers started building airplanes in a rented garage. The older one was Allan Lockheed, the younger Malcolm. Neither had an engineering degree. Their first plane was cobbled together on a borrowed Hollywood film lot — the fuselage glued from plywood. Then they broke Charles Lindbergh's long-distance flight record.
  95. 095 CVS-Health 1963, Lowell, Massachusetts. Stanley and Sidney Goldstein, two brothers, opened a small health and beauty shop with their partner — called Consumer Value Store, or CVS for short. The first store sold toothpaste, shampoo, and bandages. Then it swallowed the pharmacy next door, the clinic after that, and the insurance business after that.
  96. 096 Danaher---丹纳赫 1984. Steven and Mitchell Rales, two brothers, founded a small investment company in Massachusetts. Then they began relentlessly acquiring life sciences and industrial technology companies — using an extremely disciplined efficiency methodology called the "Danaher Business System." From water quality analyzers to mass spectrometers to single-use bioreactors — they turned the world's "measurement and diagnostic tools" into an empire.
  97. 097 Altria---奥驰亚 1847, Bond Street, London. Philip Morris opened a small shop selling hand-rolled Turkish cigarettes. Over a century later, his brand Marlboro became the most valuable tobacco brand on Earth — fused with the image of the "cowboy." Today its parent, Altria Group, is worth $116.6 billion. Meanwhile, the tobacco company's "transformation" has become one of the world's most controversial and compelling business pivots.
  98. 098 Progressive---前进保险 1937. Jack Green and Joseph Lewis founded Progressive Insurance in Cleveland. They were the first to offer "installment premiums" — letting mostly weekly-wage workers pay auto insurance weekly or monthly instead of one lump sum per year. Progressive is now one of the largest personal auto insurers in America, worth $116.1 billion.
  99. 099 Bristol-Myers-Squibb---百时美施贵宝 1887, Clinton, New York. Two young men — William McLaren Bristol and John Ripley Myers — founded the precursor to Bristol-Myers Squibb in the back room of a rented pharmacy. They didn't invent drugs — they did something stranger: they took already-effective medications on the market and recrystallized them into purer crystals, then pressed them into gelatin capsules that were easier for patients to swallow.
  100. 100 Vertex-Pharmaceuticals---福泰制药 1989, Cambridge. Joshua Boger founded Vertex Pharmaceuticals. He was 32 and a zealot for "rational drug design" — no massive screening of compound libraries by luck, but computer molecular simulation. The first decade, nearly every lead compound failed in animal or early clinical trials. Then in 1999, he wagered Vertex's entire fate on an unprecedented bet — designing molecules atom by atom to fix the underlying protein defect in cystic fibrosis.
  101. 101 Capital-One 1988. Richard Fairbank and Nigel Morris founded Capital One in Virginia. They were not traditional bankers — Fairbank was a Stanford MBA obsessed with applying statistics and experimental methods to credit cards. He ran thousands of simultaneous A/B tests — on interest rates, colors, wording, credit limits — using data to find every micro-segment of profitability the big banks ignored.
  102. 102 Quanta-Services 1997, Houston. Quanta Services started in oilfield electrical and communications cable construction. Then renewable energy and electrification exploded. Today it is the largest specialized power transmission and renewable energy engineering contractor in America — wherever a new solar farm rises, an underground high-voltage cable is laid, or an EV charging station and substation get built, Quanta's trucks are there.
  103. 103 Stryker---史赛克 1941. Dr. Homer Stryker — an orthopedic surgeon in Michigan — was driven to despair by the operating room equipment he saw. He welded a new type of fracture fixation bed in his garage using scrap stainless steel and old bed frames. He went on to design orthopedic saws, hip implants, and mobile hospital beds. Today Stryker is worth over a hundred billion dollars — any operating room on Earth likely contains a Stryker device.
  104. 104 Intuit---财捷 1983. Scott Cook sat at his Palo Alto apartment's kitchen table. His wife had covered the table with checks and bills — complaining how painful household bookkeeping was. Cook, a Harvard MBA who'd just quit P&G, picked up the phone and called banks and software companies — asking "Why is no one doing personal finance software on a PC?" Then he wrote QuickBooks.
  105. 105 Parker-Hannifin---派克汉尼汾 1917. Arthur Parker built the first reliable truck pneumatic brake fitting in a small machine shop in Cleveland. One hundred years later, Parker Hannifin is the invisible giant of motion and control technology — from commercial aircraft hydraulic systems to factory robot pneumatic valves to seals in renewable energy equipment. Market cap: over a hundred billion dollars.
  106. 106 Howmet-Aerospace 1888, Pittsburgh. Howmet's predecessor was the precision casting division of aluminum giant Alcoa. It later became independent as Howmet Aerospace. Today it is the global leader in precision castings for aerospace engines and gas turbines. Inside every Boeing and Airbus engine, inside the turbine blades that spin at temperatures above the melting point of the metal itself — those are Howmet castings.
  107. 107 Equinix 1998. Al Avery founded Equinix in Redwood City, California. His idea was strange: not selling servers, not selling network cables — selling "the place where networks shake hands with each other." Equinix data centers are the cross-connection points for every cloud provider, telecom carrier, CDN, and bank network. Today Equinix is worth nearly a hundred billion.
  108. 108 Constellation-Energy 2022. Constellation was spun off from Exelon and listed. But its lineage traces back to Baltimore Gas and Electric — founded in 1816 as one of America's earliest gas-lamp companies. Today Constellation is the largest carbon-free energy producer in the U.S. — owning dozens of nuclear power plants. In an era when AI is driving electricity demand through the roof, those nuclear reactors are suddenly worth far more than anyone predicted.
  109. 109 Southern-Company 1912. James Mitchell founded Alabama Power in Birmingham. This was the precursor to Southern Company. Today Southern is one of America's leading electric utilities — and completed America's first new nuclear reactors in decades, Vogtle Units 3 & 4, at its Vogtle nuclear plant. Vogtle spanned an entire generation from groundbreaking to commercial operation — and is now the largest clean-energy construction project in U.S. history.
  110. 110 CME-Group---芝商所 1848, Chicago. Eighty-three grain merchants gathered in a wooden building by the river and founded the Chicago Board of Trade (CBOT). They did one thing — turned wheat, corn, and oats into standardized "futures contracts" instead of haggling every batch at different qualities. From that moment, commodities became financial instruments that could be traded, hedged, and priced. Later, CBOT merged with the Chicago Mercantile Exchange to form CME Group.
  111. 111 Cadence-Design-Systems 1988, San Jose. A group of EDA engineers from Berkeley and Bell Labs founded Cadence. They made "design automation tools for chip designers" — before you draw a single transistor line, Cadence's software has already simulated tens of thousands of physical effects. Today Cadence is worth over a hundred billion dollars.
  112. 112 Adobe 1982. Two Xerox PARC engineers — John Warnock and Chuck Geschke — quit and founded Adobe in a garage. They took with them an invention Xerox refused to commercialize: PostScript — a page description language that described typefaces using mathematical curves. Then Steve Jobs commissioned them for the first Macintosh LaserWriter. Desktop publishing was born.
  113. 113 Synopsys 1986. Aart de Geus founded Synopsys in North Carolina's Research Triangle. He'd been doing chip design automation at General Electric — GE thought EDA wasn't core. He took the logic synthesis tools and made them Synopsys's foundation. Today Synopsys is worth over a hundred billion and in 2024 announced a $35 billion acquisition of Ansys — the largest semiconductor software deal in history.
  114. 114 Duke-Energy 1904. James Buchanan Duke — the American Tobacco king — founded a small hydroelectric company in North Carolina, harnessing the Catawba River. His power plants supplied textile mills and tobacco drying barns. One hundred and twenty years later, Duke Energy is one of the largest regulated electric utilities in America — serving over 8 million customers across Florida, Indiana, Ohio, and Kentucky.
  115. 115 HCA-Healthcare 1968, Nashville. Dr. Thomas Frist Sr., his son, and a physician partner founded HCA (Hospital Corporation of America). They built a modern community hospital designed to be "run by doctors themselves" — using standardized supply chains, shared services, and scale purchasing to make surgery and hospitalization cheaper and more consistently high-quality than standalone community hospitals. Today HCA is America's largest for-profit hospital system.
  116. 116 Cummins 1919, Columbus, Indiana. Clessie Cummins — a self-taught mechanic and engine fanatic — began installing diesel engines into trucks with his boss's backing. At the time, diesel engines were stationary industrial equipment — for pumping water and generating electricity. During the Great Depression in 1929, Cummins drove a truck with his homemade diesel engine across America — no refueling, no stopping — proving to a skeptical world that diesel belonged on the highway.
  117. 117 General-Dynamics 1899, Quincy, Massachusetts. An Irish-American inventor named John Holland built the U.S. Navy's first practical submarine — the Holland VI — on the banks of the Fore River Shipyard. His Electric Boat Company was the precursor to General Dynamics. Today General Dynamics is the manufacturer of the M1 Abrams main battle tank, Virginia-class nuclear submarines, and Gulfstream business jets. Market cap: nearly a hundred billion.
  118. 118 ServiceNow 2004. Fred Luddy crawled from the wreckage of Peregrine Systems — his former company had collapsed in massive accounting fraud. He built a new company in a small San Diego office: ServiceNow. Core idea: every IT workflow in an enterprise — from employee onboarding, account setup, to incident resolution — could be turned into a single automated cloud platform. Today ServiceNow is one of the world's most valuable enterprise software companies.
  119. 119 Marriott-International 1927, Washington, D.C. J. Willard Marriott and his wife Alice opened a nine-seat roadside stand — an A&W root beer shop. On hot days he saw people in cars unable to get a cold drink — so he invented the concept of "roadside cold refreshment." Later he expanded it into the Hot Shoppe restaurant chain, and then in 1957 opened his first hotel in Arlington, Virginia. Today Marriott is the world's largest hotel company.
  120. 120 KKR-&-Co. 1976. Three former Bear Stearns executives — Jerome Kohlberg, Henry Kravis, and George Roberts — founded KKR in New York. They invented a tool called the "leveraged buyout" — using borrowed money to buy undervalued public companies whole, take them private, fix the operations, and then re-list or sell them years later. In 1988, KKR pulled off the largest LBO in history: the $31.4 billion takeover of RJR Nabisco, chronicled in the book "Barbarians at the Gate."
  121. 121 Freeport-McMoRan 1912, Texas. A small mining company called Freeport Sulfur was founded. Later it discovered Grasberg in the remote mountains of New Guinea — one of the world's largest single copper and gold deposits. It then merged with McMoRan to become Freeport-McMoRan. Today it controls some of the richest copper and molybdenum reserves on Earth.
  122. 122 BNY-Mellon 1784. Alexander Hamilton founded the Bank of New York — the oldest bank in America and the first stock ever traded on the New York Stock Exchange. Over two centuries later, it merged with Mellon Financial to become BNY Mellon. Today it is one of the world's largest custodian banks — safeguarding nearly $50 trillion in assets under custody. Every mutual fund, every ETF, every sovereign bond — BNY Mellon tracks, values, and protects them.
  123. 123 Williams-Companies 1908. Dave and Miller Williams founded the Williams Companies in Arkansas — initially building oil and gas pipelines. Today it is one of America's largest interstate natural gas pipeline operators — from the Gulf of Mexico to New England, through forests, under rivers, buried underground. Roughly one-third of the natural gas that heats your home in winter and generates your electricity in summer flows through a Williams pipe.
  124. 124 Comcast 1963. Ralph Roberts bought a tiny cable TV system in Tupelo, Mississippi, with just 1,200 subscribers. That was the seed of Comcast. His son Brian Roberts later transformed that small cable company into America's largest broadband and cable media conglomerate — acquiring NBCUniversal, Universal Studios, and Sky. Today it is a media and connectivity colossus.
  125. 125 FedEx 1965, Yale University. Fred Smith proposed an "overnight express" air network model in a term paper — all packages fly to a central sorting hub, then are redistributed in the early morning to flights bound for their destinations. The professor wrote in the margin: "This idea is unfeasible. You're competing against the U.S. Postal Service and passenger airlines." Smith got a C. Six years later, he built FedEx — and proved the professor's marginal note wrong.
  126. 126 Intercontinental-Exchange 2000. Jeffrey Sprecher — a serial entrepreneur who began in power trading — founded the Intercontinental Exchange (ICE) in Atlanta. It started as an online energy trading platform. Then he made a series of audacious acquisitions — including the New York Stock Exchange itself. Today ICE is worth over a hundred billion. Nearly every futures contract, every barrel of oil traded, and every interest-rate benchmark you touch runs through ICE's systems.
  127. 127 McKesson 1833, New York City. John McKesson and Charles Olcott partnered in a pharmaceutical import and wholesale business in Lower Manhattan — importing bottles of European herbs, tinctures, and opium, selling them to American doctors and druggists. Nearly two hundred years later, McKesson is America's largest pharmaceutical distributor and healthcare supply chain company — moving every pill, every vaccine, every IV bag across the country.
  128. 128 Waste-Management 1968. Dean Buntrock founded a small garbage company in Chicago — pick up trash, bury it. Then in 1971 he merged several local haulers into Waste Management. It did something nobody thought was glamorous — bought up landfills across America and painted every truck the same shade of green. Today it is the largest waste and recycling company in North America.
  129. 129 ADP 1949, New Jersey. Henry Taub was an accountant. He noticed that every company spent enormous amounts of manual labor calculating payroll and tax withholding before each payday — a painfully inefficient, error-prone, purely manual process. He founded a company called "Automatic Data Processing," renting IBM punch-card machines to compute payroll for other companies. ADP later became the world's largest payroll and HR services provider.
  130. 130 PNC-Financial-Services 1845, Pittsburgh. Pittsburgh Trust Company opened in a small office, managing trusts and making local bank loans for the steel and railroad barons of the era. Through successive mergers — especially with National City Corporation — it became today's PNC Financial Services Group. PNC is the quintessential "super-regional bank" — deeply entrenched across the Eastern U.S. corporate and personal banking landscape.
  131. 131 Monster-Beverage 1935, California. A family juice operation founded by Hubert Hansen began selling fresh juice to Southern California movie studios. Decades later, his grandsons Rodney Sacks and Hilton Schlosberg took a bland energy drink brand called Monster — and transformed it into a feral black-and-green-clawed beast. Today Monster Beverage is worth tens of billions and is the second-largest energy drink brand on the planet.
  132. 132 Fortinet 2000, Sunnyvale. Ken Xie — a Tsinghua-educated cybersecurity engineer — founded Fortinet. His previous company NetScreen had been acquired by Juniper; he didn't retire. He built another, cheaper next-generation firewall — accelerated by in-house custom ASIC chips. Today Fortinet is worth hundreds of billions.
  133. 133 U.S.-Bancorp 1863, the very month Lincoln signed the National Banking Act, the First National Bank of the United States was founded in Minneapolis — the ancestor of today's U.S. Bancorp. One hundred and sixty years later, U.S. Bancorp has thousands of branches across the Midwest and West, plus trillions in trust and payment operations. It is one of the five largest banks in America by assets.
  134. 134 Elevance-Health 1944, Indianapolis. Blue Cross Blue Shield of Indiana was founded as Mutual Medical Insurance — the precursor to Anthem, later renamed Elevance Health. Today it is one of the largest health insurers in America, providing coverage to over 40 million members.
  135. 135 UPS 1907, Seattle. A 19-year-old named Jim Casey used a borrowed $100 and a secondhand bicycle to start a local messenger company called "American Messenger." He personally pedaled through the rain from one office building to another, delivering letters and parcels. Today UPS is one of the world's largest package delivery and logistics companies — that bicycle has become an armada of brown trucks and cargo jets spanning the globe.
  136. 136 American-Tower 1995, Boston. American Tower was spun off from American Radio Systems (a predecessor of CBS). The logic was simple: broadcasters and mobile carriers were frantically mounting antennas on rooftops and water towers — why not turn "the antenna pole itself" into independent, rentable infrastructure? Today American Tower owns over 200,000 communications towers worldwide — every cell phone call you make likely touches one of its steel towers.
  137. 137 CSX-Corporation 1827, Maryland. The Baltimore and Ohio Railroad — one of America's oldest freight and coal-hauling railroads — laid its first cornerstone in the presence of Charles Carroll, a signer of the Declaration of Independence. Today CSX operates tens of thousands of miles of freight rail across the Eastern U.S. — moving Appalachian coal, Midwestern grain, Gulf Coast chemicals, and East Coast container cargo.
  138. 138 Enterprise-Products 1968, Houston. Dan Duncan launched Enterprise Products with a single propane truck and a small propane transfer station. Then he laid propane pipelines into, out of, and across the Gulf Coast. Today Enterprise Products Partners is one of the largest midstream energy pipeline and export terminal companies in North America — moving crude oil, natural gas liquids, chemicals, and refined products across the continent.
  139. 139 Schlumberger 1926, Alsace, France. Conrad and Marcel Schlumberger — two geophysicist brothers — invented the first downhole resistivity logging cable for oil exploration. They lowered electrodes down a wellbore, sent electric current into the rock formations, and measured the return waveform — thereby inferring what lay dozens of meters beneath without ever drilling a core sample. Today SLB is the world's largest oilfield services company.
  140. 140 SLB-Schlumberger 1912, Alsace, France. Two Schlumberger brothers — Conrad and Marcel — rigged a strange device in their family's backyard barn: a wooden crate with four copper electrodes, battery-powered, buried in the dirt. They weren't looking for groundwater — they were measuring variations in soil resistivity. Conrad was a physicist, Marcel a coal-mine engineer. Together they invented the science of subsurface imaging that would one day find oil fields across the planet.
  141. 141 Airbnb October 2007, San Francisco. Two Rhode Island School of Design graduates — Brian Chesky and Joe Gebbia — couldn't make rent. They noticed a design conference had booked every hotel in the city solid. They pulled out three air mattresses, laid them in their living room, and posted a message: "Airbed + lodging + breakfast — $80." Ten years later, Airbnb is one of the largest hospitality platforms on Earth — without owning a single hotel.
  142. 142 Boston-Scientific 1979, Boston. Two medical device salesmen — John Abele and Pete Nicholas — sat in a Dunkin' Donuts next to Massachusetts General Hospital. Abele had just quit another interventional device company that rejected his coronary balloon catheter design. He drew a sketch on a napkin: an ultra-thin balloon at the tip of a catheter that could travel through blood vessels and be inflated inside a blocked coronary artery. That napkin became Boston Scientific.
  143. 143 Moodys 1900, Lower Manhattan. A young man who'd spent years as a Wall Street newsboy and securities statistician — John Moody — sat in a tiny unheated office drawing grids. He was building something every Wall Street bank considered "completely worthless": a single book organizing the balance sheets, earnings, and management quality of every American railroad company. That book became Moody's Manual — the ancestor of Moody's Investors Service.
  144. 144 Mondelez 1898, Chicago. A small grocery wholesaler named Fred E. Stephen bought the exclusive U.S. distribution rights for Toblerone Swiss triangular chocolate from a traveling salesman just back from Switzerland. Then he added: Milka, Lu, and Nabisco (Oreo's parent company, acquired by Kraft in 1999). Today Mondelez International is the world's largest chocolate and biscuit snacking empire.
  145. 145 Ciena 1992, Maryland. Three engineers from General Instrument's fiber-optic and military communications division — David Huber, Kevin Hayes, and Patrick Nettles — quit their jobs and raised their first $2 million from a venture capital fund. Their idea sounded insane at the time: use 16 different wavelengths of laser on a single strand of optical fiber and turn each wavelength into an independent data channel. Today that technology is called DWDM — and it powers the entire global internet backbone.
  146. 146 Marsh-McLennan 1871, Chicago. Henry W. Marsh made a decision after the Great Chicago Fire burned down half the city — he would no longer be just a railroad insurance broker. He would build a firm that "puts all industrial risks — from fire to boiler explosion to marine — onto a single sheet of paper and prices the premium." That was the origin of Marsh. In 1905, Alexander McLennan independently founded McLennan. The two merged and became Marsh & McLennan, today the world's largest insurance broker.
  147. 147 Bloom-Energy 2001, NASA Ames Research Center. An Indian-American scientist named Dr. K.R. Sridhar had just designed a solid-oxide fuel cell device for NASA that could produce oxygen from the carbon dioxide and hydrogen of the Martian atmosphere. The device ran successfully in simulated Mars conditions. When Sridhar returned to Earth, he ran that Martian chemistry in reverse — feeding it natural gas and air to generate electricity — and called the company Bloom Energy.
  148. 148 Northrop-Grumman 1930, Inglewood, California. A Norwegian-American engineer who'd walked out of a Los Angeles aircraft factory built a shed-style plant on the edge of the desert. His name was Jack Northrop. His favorite plane — the Alpha — pioneered the all-aluminum monocoque fuselage and full-metal stressed-skin design, while every other plane was still biplane canvas and wood. He made the fuselage as smooth as a polished mirror — and it flew faster than anything in the sky.
  149. 149 Simon-Property-Group 1960, Indianapolis. A young man named Melvin Simon — just discharged from the Army — used his Korean War savings and joined his brother Herb to build their first open-air shopping plaza on a cornfield outside Indianapolis. He wasn't a real estate developer — he was a mortgage broker. His logic was beautifully simple: if we build a mall, the shops will come, and we'll finance their leases.
  150. 150 Monolithic-Power-Systems 1997, Los Gatos, California. A former analog power chip engineer from TI and ADI — Michael Hsing — began designing a new DC-DC converter chip at his dining room table. At the time, analog power chips were a fortress dominated by TI, Linear Tech, Maxim, and Power Integrations — and he was one man with a soldering iron in a suburban California house.
  151. 151 Cigna 1792, Philadelphia. The United States' first marine insurance company — Insurance Company of North America (INA) — was chartered by the Pennsylvania legislature. Its first president was Dr. Benjamin Rush, a signer of the Declaration of Independence. But its true core was what it insured — from slave-ship cargo losses in the 1790s to 20th-century corporate health plans. Cigna today is the descendant of that history — one of America's largest health insurers.
  152. 152 Sherwin-Williams 1866, Cleveland. Henry Sherwin — the son of a bookseller from rural Ohio — bought a bankrupt paint store's inventory and formula notes for $2,000. He opened every paint barrel and measured the ingredient ratios of each batch — discovering that three batches of the same white paint from the same supplier varied in titanium dioxide content by 40%. He decided the world needed a paint company that actually knew what was inside its own cans.
  153. 153 Lumentum 2015, Milpitas, California. An optical networking equipment company called JDSU — after nearly two decades of struggling — performed a surgical split: it spun off the optical communication laser and photonics business as Lumentum, and the network test and measurement business became Viavi. The day the separation papers were signed, Lumentum became a pure-play photonics company — supplying the lasers and optical components inside every data center and fiber-optic network.
  154. 154 OReilly-Automotive 1957, Springfield, Missouri. Charles F. O'Reilly and his son Charles H. O'Reilly — a World War II auto mechanic — rented an abandoned gas station shed off Route 66 and opened the first O'Reilly Auto Parts store. On opening day, their shelves held exactly two things: spark plugs and radiator hoses.
  155. 155 Emerson 1890, St. Louis, Missouri. Judge John Wesley Emerson — grandson of a Scottish-American immigrant — opened a small electric motor factory on the banks of the Mississippi River. He made only one thing: electric motors that ran on the AC current from Niagara Falls. At the time, Edison's DC grid was locked in a war with the Westinghouse-Tesla AC grid for the future of electricity. Emerson bet on AC — and won.
  156. 156 Regeneron 1988, Tarrytown, New York. A young neuroscientist named Leonard Schleifer — who'd spent years at Cornell Medical Center studying nerve growth factor — locked himself in a rented Westchester County laboratory. He and his chief scientist, George Yancopoulos (a Columbia immunology postdoc), pooled $1 million and decided to build a company that made fully human monoclonal antibodies — from mice genetically engineered to carry human immune-system genes.
  157. 157 Apollo-Global-Management 1989, New York. Three young men from Drexel Burnham Lambert (Michael Milken's junk bond empire) — Leon Black, Josh Harris, and Mark Rowan — rented a small office to pursue a business "basically nobody dared to do head-on": using junk bond financing to acquire distressed or out-of-favor companies that other funds wouldn't touch. That firm became Apollo Global Management, today one of the largest alternative asset managers on Earth.
  158. 158 Marathon-Petroleum 1887, northwestern Ohio — near the village of Findlay. A small crude oil operator named John Vandervis poked the region's first commercial oil well in the middle of a cornfield. When the oil gushed, he had no pipeline, no refinery, no storage tanks. He loaded wooden barrels onto a farmer's wagon and used horses to haul the crude to the nearest railroad siding, one barrel at a time.
  159. 159 3M 1902, Two Harbors, Minnesota. Five local businessmen — a doctor, a lawyer, two grocers, and a railroad ticket agent — signed a partnership agreement. They bought the mining rights to a corundum deposit on the north shore of Lake Superior, believing crushed corundum could be the abrasive for sandpaper and grinding wheels. On the first day at the mine, the rock wasn't corundum at all. It was worthless low-grade anorthosite. That failure forced them to invent something entirely new — and that something became 3M.
  160. 160 Valero-Energy 1980, San Antonio, Texas. The CEO of a pipeline and natural gas company called Coastal States Gas Corporation — Oscar Wyatt — spun off its downstream refining assets to shareholders and created a new company, naming it Valero. The name came from the Mission San Antonio de Valero — the Alamo. Valero would become the largest independent refiner in the United States.
  161. 161 Illinois-Tool-Works 1912, Chicago. Byron C. Smith — a German-American machinist — opened his first workshop near the Union Stockyards. His workshop did only one thing: built better, more precise gear-drive systems and machine-bed slides that didn't fall off their chains — for Chicago's exploding meatpacking, printing, and metal-stamping factories. Today Illinois Tool Works is a deeply decentralized industrial conglomerate worth over a hundred billion.
  162. 162 Kinder-Morgan 1997, Houston. Richard Kinder — who'd spent twenty years as COO and Vice Chairman at Enron under Ken Lay — left Enron a year after falling out with Lay and began doing what he'd always pushed at Enron but Lay had vetoed: using low leverage to buy sleeping assets — pipelines. Kinder bought a few narrow old Texas intrastate natural gas liquid pipelines from a bankruptcy auction. Today Kinder Morgan is one of the largest energy infrastructure companies in North America.
  163. 163 Hilton 1919, Cisco, Texas. A young man named Conrad Hilton tried to buy a small bank in Cisco. The deal collapsed. Frustrated, he walked into the Mobley Hotel across from the bank — a small inn overflowing with oil workers, so crowded that beds were rented in eight-hour shifts. He bought the hotel that very night. That became the first Hilton.
  164. 164 Coherent 1966, Palo Alto, California. A handful of young graduates from Stanford's electrical engineering and physics departments — led by Eugene Watson — rented a small industrial bay under 500 square feet in the Stanford Industrial Park and began manufacturing industrial carbon dioxide (CO2) lasers. Before that, lasers were almost entirely laboratory devices — locked on optical tables. Coherent turned lasers into industrial tools.
  165. 165 American-Electric-Power 1906, Columbus, Ohio. Richard E. Breed — an electrical engineer who'd come from Chicago Edison — merged dozens of scattered village and small-town early power plants across Ohio into a single holding company, initially named American Gas and Electric, later renamed American Electric Power (AEP). At the time, each town's power plant was a standalone unit; Breed connected them into the first major interstate electric grid.
  166. 166 EOG-Resources 1999, Houston. Enron Oil & Gas (EOG) — the "boring" oil exploration subsidiary of Enron — was spun off and listed independently by Ken Lay's order, so that the assets released from it could be funneled into Enron's broadband and power trading empire. The day the spinoff closed, EOG Resources' management sighed with relief. They had just been cut loose from the largest corporate fraud in American history — before anyone knew it.
  167. 167 Datadog 2010, New York City. Two French software engineers — Olivier Pomel and Alexis Le-Quoc — stared at server monitoring dashboards in a cramped Manhattan shared office. They'd just left Wireless Generation (a New York education data company) because they'd both seen the same problem there: monitoring cloud infrastructure involved gluing together a dozen half-working open-source tools. So they built a single unified monitoring platform — Datadog.
  168. 168 Phillips-66 2012, Houston. ConocoPhillips decided to spin off its downstream businesses — refining, chemicals, pipelines, and gas stations — into an independent company. The new company's name was pulled from ConocoPhillips history: Phillips 66. Originally just a gasoline blend brand name launched in 1927, it referenced the number 66 in an early road test where the fuel outperformed competitors on U.S. Route 66.
  169. 169 Ecolab 1923, St. Paul, Minnesota. A chemist named Merritt J. Osborn — who'd previously researched dairy sterilization at the U.S. Department of Agriculture — pitched a "solid dishwasher descaler" to a hotel manager in a downtown hotel banquet hall. At the time, many hotel kitchens washed dishes with water that wasn't hot enough, leading to frequent foodborne illnesses. Osborn's product fixed that. That was the birth of Ecolab, today the global leader in cleaning and sanitation.
  170. 170 Colgate-Palmolive 1806, Dutch Street, New York City. William Colgate — the 23-year-old son of an English immigrant — opened a small workshop beneath a clockmaker's shop. He made just three products: soap, candles, and starch. His reasoning: all three required the same base raw material — animal fat. He sold his goods from a pushcart at the New York Harbor docks to ships about to sail.
  171. 171 Norfolk-Southern 1830, Charleston, South Carolina. The Charleston City Council — lobbied by local merchants — approved the issuance of the region's first railroad bonds to finance a rail line from Charleston to Hamburg (a pure cotton transport route). The Charleston-Hamburg Railroad was among the earliest ancestors of the Southern Railway system. Later, in 1838, another ancestor was born in Norfolk, Virginia. The two systems eventually merged into Norfolk Southern.
  172. 172 Digital-Realty 2004, San Francisco. A real estate investor named Michael Foust, who'd successfully managed data center assets for GI Partners (a mid-market private equity fund), decided to spin those data center assets into a standalone public company. He named it "Digital Realty Trust." Today Digital Realty is one of the largest data center REITs in the world — it owns the physical buildings where the cloud lives.
  173. 173 Royal-Caribbean 1968, Oslo, Norway. Three Norwegian shipowners — Arne Wilhelmsen, Sigvald Eivindson, and Anders Wilhelmsen — sat in their shipbroker office staring at the collapsing transatlantic passenger liner data and said: "Nobody wants to cross the Atlantic by ship anymore — a plane does it in seven hours. We have to build something in the sea that an airplane cannot: a floating vacation."
  174. 174 Robinhood 2013, Palo Alto, California. Two Stanford math and physics graduates — Vladimir Tenev and Baiju Bhatt — sat in a coffee shop surrounded by dozens of printed pages of SEC regulations. They'd discovered something: the underlying execution cost of U.S. stock trades had already fallen to zero — but every traditional broker still charged per-share or per-trade commissions. They built an app with zero commissions.
  175. 175 Ross-Stores 1950, San Bruno, California. A longtime San Francisco department store buyer named Morris "Morrie" Ross noticed a strange truth: every major department store and brand boutique had an "inventory black hole" — last season's unsold clothes, odd-size shoes, and returned cosmetics with damaged packaging. These goods were written off as losses. Ross offered to buy them at a deep discount and resell them for a fraction of the original price. Today Ross Stores is America's largest off-price retailer.
  176. 176 Comfort-Systems-USA 1997, Houston. A mid-sized HVAC (heating, ventilation, and air conditioning) contractor — formed by engineers and project managers who'd splintered from larger mechanical contracting firms — went public on Nasdaq. It acquired dozens of local HVAC, electrical, and plumbing contracting firms and integrated them under a single operating platform. Today Comfort Systems USA is one of America's largest specialty mechanical contractors.
  177. 177 DoorDash 2013, a Stanford University classroom. Four Chinese and Chinese-American students — Tony Xu, Andy Fang, Stanley Tang, and Evan Moore — were researching a small class project on "digitization for local small businesses in Palo Alto." They walked into a dessert shop. The owner flipped open a stack of paper on the front counter. The stack was all the delivery orders they couldn't fulfill because they had no driver. The students built a prototype delivery platform. Today DoorDash is America's largest food delivery company.
  178. 178 General-Motors 1908, Flint, Michigan. William C. Durant — a self-taught businessman who'd become one of Michigan's richest men making horse-drawn carriages in the late 19th century — bundled Buick, Oldsmobile, and Cadillac together into a new company called General Motors. His core logic wasn't building cars — it was using a parent holding company to own multiple car brands that could share parts, engineering, and distribution while competing against each other.
  179. 179 Warner-Bros-Discovery 1923, Hollywood. Four Polish Jewish immigrant brothers — Harry, Albert, Sam, and Jack Warner — rented an empty meat cold-storage warehouse on Gower Street and turned it into a film screening room and office. They had one old camera and one horse that would perform in front of the lens — Warner Bros.' first films were dogs and horse acts. In 1927, they gambled everything on "talkies" with The Jazz Singer — and changed cinema forever.
  180. 180 Energy-Transfer 1996, Dallas, Texas. Kelcy Warren — an engineer who'd run a small pipeline welding and natural gas processing plant in Texas — started with just $20 million and bought a narrow, almost-defunct old intrastate Texas natural gas liquids pipeline from a bankruptcy auction. Warren said to his partner: "If we can just get one more gallon moving through this pipe every minute, we survive." Today Energy Transfer is one of the largest pipeline empires in America.
  181. 181 Rocket-Lab 2006, Auckland, New Zealand. A New Zealander with no university degree — Peter Beck — began hand-mixing rocket fuel and building small solid rockets in a corrugated-iron shed in his backyard. No NASA. No government contracts. No space agency. All he had was an old lathe his mother didn't want and a secondhand mill. With these he built his first rocket engine. Today Rocket Lab is the second-most-frequent orbital launch provider on the planet.
  182. 182 Cloudflare 2009, San Francisco. Matthew Prince — Harvard Business School and Harvard Law graduate — sat in a San Francisco bar with his former Harvard computer science classmates Lee Holloway and Michelle Zatlyn, trying to solve a project they'd worked on for years: Project Honey Pot (a project tracking email spammers). They realized the same technology that tracked spammers could also protect websites. Today Cloudflare is the shield in front of a huge swath of the internet.
  183. 183 Air-Products 1940, Detroit. A young Ohio chemical engineer named Leonard Parker Pool — after earning his chemical engineering PhD from the University of Michigan — spent several years as a salesman and plant manager at a Detroit industrial gas company (Burdett Oxygen). He noticed something: all the factories that used oxygen, nitrogen, and acetylene for welding and cutting had to buy their own gas generators. He had a radical idea: build enormous centralized gas plants and pipe the gas directly into customers' factories.
  184. 184 Cintas 1929, Cincinnati. Richard T. Farmer — a former Acme Paper sales employee laid off during the Great Depression — salvaged hundreds of returned and slightly damaged industrial rags from Acme's scrap warehouse. He washed them, re-dyed them, and drove his own old car to sell these "reborn rags" to Cincinnati factories and repair shops. That was the birth of Cintas, today one of America's largest uniform and workplace services companies.
  185. 185 TransDigm 1993, Cleveland. A finance and M&A lawyer named Wally Nichols — who'd spent years in venture capital and private equity — led a small team to buy the aircraft brake components division of Kelsey-Hayes, a forgotten manufacturer of aerospace brake actuators and hydraulic parts inside a large conglomerate. Nichols discovered the division made one thing: proprietary, flight-critical parts that were the only approved replacement component on specific aircraft types. Today TransDigm is one of the most profitable aerospace companies in the world.
  186. 186 Motorola-Solutions 1928, Chicago. Paul V. Galvin — a former battery salesman who'd lost everything when his company went bankrupt — rented a small workshop on West Harrison Street in downtown Chicago. He and five employees began installing battery eliminators and radios into large wooden boxes on a conveyor belt. Galvin hung a sign in the window: "Galvin Manufacturing Corporation." Later he renamed the company Motorola — after the "motor" in motorcar and the "ola" of Victrola.
  187. 187 MicroStrategy 1989, Wilmington, Delaware. An MIT engineering graduate who'd done biostatistics and numerical simulation at DuPont — Michael Saylor — teamed up with his MIT fraternity brother Sanju Bansal to start a company. Its first business was not "buying Bitcoin" — it was business intelligence (BI) software. Then, in 2020, Saylor did something no public company CEO had ever attempted: he put the company's entire treasury reserves into Bitcoin. Today MicroStrategy holds more Bitcoin than any other corporation.
  188. 188 The-Travelers-Companies 1864, Hartford, Connecticut. James G. Batterson — a former stonecutter and railroad ticket agent — hung a new sign outside his downtown Hartford office. Two months earlier, he'd first encountered the term "accident insurance" outside the Bank of England — the idea of insuring people against everyday mishaps, not just maritime catastrophe. He brought that idea to America.
  189. 189 Baker-Hughes 1907, Navasota, Texas — an oil-patch town on the Gulf Coast. A machinist named Reuben C. Baker, who'd maintained steam engines and mud pumps for local drilling contractors, invented an extremely simple small tool in his own workshop: a casing shoe that could be lowered into a wellbore on a cable, then slide and seal inside oil well casing. That tool became the seed of Baker Hughes, today one of the world's largest oilfield services companies.
  190. 190 Republic-Services 1980s, Phoenix, Arizona. Wayne Huizenga — a serial entrepreneur who would later create Blockbuster Video and co-found Waste Management — merged a string of independent local garbage haulers into a company called Republic Waste Services. Today Republic Services is the second-largest waste and recycling company in the United States.
  191. 191 Nike 1964. A Stanford MBA graduate and college track runner borrowed $50 from his father and drove his green Plymouth Valiant to every high school track meet in Oregon. His trunk was full of Japanese running shoes — Onitsuka Tigers. After the races, he'd walk up to coaches and kids and say: "Try these." His name was Phil Knight. Today Nike is one of the most valuable brands on Earth.
  192. 192 Keysight 1939, Palo Alto, California — right next to Stanford. Bill Hewlett and Dave Packard baked their first audio oscillator — the HP200A — using a secondhand Sears electric oven, and sold it to Walt Disney Studios for testing the surround-sound system for Fantasia. Keysight Technologies is the direct descendant of that garage-born measurement instrumentation business, spun off from HP in the 21st century.
  193. 193 Sempra 1998, San Diego, California. In the wave of California electricity market deregulation, a local natural gas utility named Enova — whose roots trace back to the San Diego Gas Light Company founded in 1881 — merged with Southern California Gas Company to form Sempra Energy. Today Sempra is one of the largest energy infrastructure and utility holding companies in North America.
  194. 194 Carvana 2012, Phoenix. Ernie Garcia III — just a few years out of Stanford Law School — worked in sales operations at his father Ernie Garcia II's DriveTime used-car finance empire. He noticed something: any car on a physical auction lot couldn't be sold online because there were no high-quality photos and no way to guarantee condition. So he built a system to photograph every car in 360 degrees, certify it, and let people buy a used car entirely on their phone. Today Carvana is the fastest-growing used-car retailer in America — and its multi-story glass car vending machines are visible from highways across the country.
  195. 195 United-Rentals 1997, Greenwich, Connecticut. Bradley Jacobs — a young man who'd done M&A at Waste Management — decided to leave the garbage business and enter a field that wasn't even considered a real industry: equipment rental. The American construction, industrial, and emergency repair equipment market was a fragmented patchwork of thousands of small, locally owned shops. He began buying them one by one. Today United Rentals is the largest equipment rental company on Earth.
  196. 196 Truist-Financial 2019, the American Southeast. SunTrust Banks (Atlanta, dating back to a commercial bank trust company founded in 1888) and BB&T (Winston-Salem, North Carolina, dating back to a rural bank founded in 1872 that made harvest loans to farmers) — merged to form Truist Financial. Overnight, Truist became one of the largest regional banks in America.
  197. 197 Paccar 1905, Seattle, Washington. William Pigott — a mechanic who'd repaired small steam locomotives on logging and coal-mining rail lines in the Pacific Northwest — partnered with a timber merchant to found the Seattle Car Manufacturing Company, building their first steam locomotive and later their first "iron truck chassis." They soon renamed the company Pacific Car and Foundry Company — Paccar. Today Paccar manufactures Kenworth and Peterbilt trucks sold worldwide.
  198. 198 Aflac 1955, Columbus, Georgia. John B. Amos — a longtime local life and health insurance broker — joined his two brothers, Paul S. Amos and Bill Amos, to hang a new sign on a small office door in Columbus: American Family Life Assurance Company — AFLAC. Their first product: supplemental cancer insurance that paid cash directly to the patient upon diagnosis.
  199. 199 Realty-Income 1969, Maryland. William E. Clark — a private real estate investor who owned a few small retail properties in Southern California — and several partners decided to form an extremely unusual REIT: they would buy only one type of property — freestanding single-tenant retail and restaurant buildings — and sign ultra-long "triple net lease" (NNN) contracts, where the tenant pays all taxes, insurance, and maintenance. Realty Income trademarked itself as "The Monthly Dividend Company."
  200. 200 WW-Grainger 1927, Chicago. William W. Grainger — a young manager at a Chicago electrical wholesale distributor — noticed something odd. When his wholesale company sold a generator to a factory, the customer would use the same purchase order to ask: "Do you also carry that special copper bus bar between the generator and the switchboard?" Big companies didn't carry those small but essential parts. Grainger decided to build a catalog of every single one of them — and W.W. Grainger became the largest industrial supply company in America.
Chapter 1 NVIDIA

32 Years, $5 Trillion — He Did Only One Thing: Bet on a Future Nobody Believed In

Not funding.
Not acquisitions.
Not government subsidies.
Not riding a trend.
Just one man, in a time when everyone said "this is useless," persisting for 18 years.
18 years later, he built his company into the most valuable enterprise on earth: $5.27 trillion.
Let me tell his story first. After that, we'll talk about why 99% of founders bet their time on the wrong things.

In a booth at Denny's, three men ordered coffee and french fries.
April 1993, San Jose.
Jensen Huang sat by the window, across from Chris Malachowsky and Curtis Priem.
Spread across the table were napkins covered in circuit sketches.
They were debating a question nobody could answer —
Beyond gaming, what else could a GPU do?
Jensen Huang said: everything.
Chris laughed. Curtis stayed silent.
Later, they walked out of Denny's carrying three things: $200 in startup capital, a few napkins covered in sketches, and an idea nobody believed in.
That year, Jensen Huang was 30 years old.

Stop here for a moment. I need to make clear what it is he built.

GPU — Graphics Processing Unit.
Sounds like a graphics card. You install it in a computer. You play games.
If you had said that in 1995, nobody would have disagreed.
The entire Silicon Valley saw the GPU as a narrow consumer-electronics accessory — there was a market, but the ceiling was right there, obvious to everyone.
Jensen Huang saw it differently.
He believed the GPU did more than render game graphics — it could do mathematical computation.
Massively parallel computation.
People thought he was talking nonsense. Because saying "parallel computing" in 1995 was like saying "air can fly" in 1880.
But Jensen Huang kept turning the question over: if a GPU can handle thousands of computational tasks simultaneously, why should it only be used for drawing pictures?
He turned that question over for 13 years before making his move.
But first, let me cover 1993 to 1996.
In those three years, he came within a hair's breadth of killing the company entirely.

In 1995, NVIDIA released its first product — the NV1.
Two years of development. Two rounds of funding burned through.
The result?
A disaster.
The NV1 used a proprietary rendering technology that was incompatible with the entire industry.
Sega had placed a large order, but canceled it after seeing the finished product.
Sales approached zero.
Spring 1996. Jensen Huang sat in his office, staring at the layoff list.
He had to cut half his workforce.
From 100 people down to 50.
That night, he sat alone in his office until two in the morning. The next day, he called employees into the conference room, one by one, and told them himself.
He later described that moment of self-doubt in an interview:
"What if I was wrong from the start?"
"What if the GPU really is only good for gaming?"
"What if I've led 100 people to the edge of a cliff, and there's nothing on the other side — how do I bear that responsibility?"
In the end, he made two decisions.
The first: pivot entirely to the Direct3D standard and abandon the proprietary approach.
The second: ask Sequoia Capital for another $2 million.
During the investment committee meeting, the Sequoia partner said a line that has since been quoted countless times:
"We're not investing in a company. We're investing in a person."
April 1997. The RIVA 128 launched.
At Comdex, the line of people waiting to try it stretched from the booth all the way to the exhibition hall entrance.
Four months. One million units sold.
NVIDIA had survived.
But that's not what made Jensen Huang great.
It was what came next.

In 2006, Jensen Huang brought a new idea to the board.
It was called CUDA.
In plain terms — make the GPU no longer just a graphics card, but a general-purpose computing platform.
Scientists could use it to compute protein folding.
Engineers could use it to simulate fluid dynamics.
AI researchers could use it to train neural networks.
Nobody on the board understood what he was talking about.
Investors started calling in: had the money gone to his head?
CUDA's annual R&D investment exceeded $500 million, and the revenue source at the time was —
Zero.
No one needed GPUs for computation. No one at all.
Year one, CUDA's adoption rate was so low it barely registered.
Year two.
Year three.
Year five.
All the way until 2012.
A graduate student named Alex Krizhevsky used two NVIDIA GPUs to train a deep neural network for image recognition.
Its recognition accuracy crushed every traditional algorithm.
AlexNet was born. The deep learning era had begun.
Suddenly, everyone working in AI realized —
The only hardware they needed was called the NVIDIA GPU.
From 2006 to 2012. He had waited six full years.
In 2016, OpenAI received its first NVIDIA DGX-1 supercomputer. Jensen Huang delivered it personally.
He signed the chassis:
"To Elon and the OpenAI team — to the future of computing."
Seven years later, ChatGPT ran on tens of thousands of NVIDIA GPUs.
Another year later, NVIDIA's market cap crossed $2 trillion.
In 2025, it crossed $5 trillion.
The seed planted in 2006 had taken 18 years to grow into the single largest tree in the history of human commerce.

Now stop for a moment.
Reading this far, I want to ask you a question.
How many years have you been building your company?
Have you ever committed to a direction for more than five years — even when there was zero return, when everyone was questioning you?
Not just enduring and waiting. I mean actively investing, steadily doubling down.
If your answer is zero —
Don't rush ahead. Sit with that question first.

I've studied Jensen Huang for a long time.
Here's the plain truth. It comes down to three things. Any founder, in any industry, can replicate them.
First: he didn't chase trends — he waited for the trends to come to him.
I've seen too many founders jump on whatever is hot.
Web3 in 2021. AIGC in 2022. Embodied intelligence in 2023.
A new direction every year, starting from zero every year.
Jensen Huang did the opposite.
He poured investment into CUDA for 18 years. The first six years: completely no return. Year seven: AI arrived.
He wasn't chasing the wind. He was squatting on the road the wind had no choice but to take.
The direction you're on right now — how many years are you willing to squat there?
Second: he didn't do addition. He did multiplication.
CUDA wasn't a new product. It was a layer of infrastructure that made every NVIDIA GPU more valuable.
With CUDA, the GPU went from "gaming accessory" to "computing platform."
Once a platform is adopted, the switching cost is staggeringly high.
Millions of AI engineers worldwide write code that runs on the CUDA architecture.
You want to switch to another GPU vendor? Go ahead. But you'll have to rewrite all your code.
That's the moat — not patents, not pricing, but the ingrained habits of millions of people.
Does your current product have a layer of infrastructure that makes it impossible for customers to walk away?
Third: he turned "technical judgment" — the most intangible capability — into the most concrete competitive barrier.
Jensen Huang once said:
"What I see is not today's market. What I see is that three to five years from now, the GPU will become the core of every computer."
He said something similar in 1999. He said it again in 2006. He said it again in 2015.
Every time, people laughed at him.
But across 30 years, he has never been wrong.
This isn't luck. This is the judgment he accumulated through 30 years of deep cultivation in one direction — GPUs.
That kind of judgment can't be bought with money. Investors can't hand it to you. It can only be earned through time and focus.

So here's the question.
What domain is your judgment in right now?
How many years have you been cultivating it deeply?
Can you see what your customers will need three years from now?
If your industry gets hit by a tsunami like the one AI brought to GPUs —
Can you, like Jensen Huang, be the first to see the waveform on the seismograph?
Those 18 years of CUDA investment — Jensen Huang fired his own bullets to make them happen.
He used 18 years to prove one thing:
Long-term thinking isn't waiting. It's acting earlier than everyone else — and then waiting longer than everyone else.

Contents Next →
Chapter 2 Google

26 Years, $4.6 Trillion — They Did Only One Thing: Refused to Sell a Search Engine Nobody Wanted to Buy

Not burning cash for growth.
Not building distribution channels.
Not constructing an ecosystem.
Not IPO hype.
It was this: at the moment when every buyer came knocking, two Stanford PhD students said "no."
And then they spent 26 years turning the thing nobody wanted into a $4.6 trillion empire.
Let me tell their story first. After that, we'll talk about why most founders sell their companies far too early.

Summer 1998, Palo Alto. An office carved out of a garage.
Sergey Brin stood on a chair, pressing acoustic foam onto the ceiling.
Larry Page lay on the floor, shoving a server built from Lego bricks into the corner.
The garage had no air conditioning. Bay Area, August. 95 degrees.
They had just been kicked out of their Stanford dorm.
Page's advisor had said one thing:
"If you keep wasting time on that search engine of yours, you will never earn your PhD."
Page glanced at Brin.
Brin shrugged.
Both of them bent back down and kept building with Lego.
The thing was called BackRub — later renamed Google.

Stop here for a moment. I need to make clear what it is they built.

A search engine. Sounds unremarkable.
By 1998, the internet already had dozens of search engines. Yahoo, AltaVista, Excite, Lycos, HotBot.
Every single one was bigger than Google. Every single one had more money.
Trying to build a new search engine in that market was like trying to build a new WeChat today — everyone would think you were insane.
But Page had discovered something while working on his Stanford thesis.
Every existing search engine ranked results by how many times a keyword appeared.
Whoever stuffed "computer computer computer" onto their page the most times ranked highest.
This was absurd.
A website's quality doesn't depend on how many times it repeats the word "computer."
It depends on how many other websites are willing to link to it, cite it, recommend it.
Just like academic papers — a paper's value isn't measured by how often it uses the field's terminology, but by how many people cite it.
Page turned this insight into the PageRank algorithm.
For the first time in human history, the internet's own voting mechanism decided the ranking of search results.

Then came 1999.
That year, Google's daily search volume exploded from 10,000 queries to 500,000.
Page and Brin were happy for a week.
Then they discovered a terrifying truth —
They were out of money.
Couldn't afford servers. Couldn't afford rent. Couldn't even afford food.
At two in the morning, Page looked up from his code and saw Brin prying quarters out of the vending machine coin return to buy a pack of instant noodles.
That day, the two of them made a decision.
Sell Google.
They brought an offer to Excite's CEO, George Bell.
$750,000.
A company processing 500,000 searches a day, asking price: $750,000.
George Bell thought about it for three seconds.
"Too expensive."
He wouldn't buy it.
He later became the most ridiculed man in business school case studies.
But in 1999, no one thought he was stupid. Back then, $750,000 for a search engine really was expensive.
The night they were rejected, Page and Brin returned to the garage.
Page sat beside the Lego server, silent for a long time.
Brin said one thing —
"Since nobody wants it, let's just build it into a company ourselves."
The two of them took apart their servers and rebuilt them, over and over.
They weren't saying "let's start a company."
They had been backed into a corner, with only one road left.
That road was called — do it yourself.

  1. The peak of the dot-com bubble's madness.
    Google signed search outsourcing deals with AOL and Netscape, multiplying their traffic tenfold overnight.
    But investors began to panic.
    "These two kids won't even wear suits." "They need adults to manage them."
    The board summoned them to a meeting.
    "You must find a CEO. A real CEO."
    Page asked: why?
    "Because the two of you don't understand management."
    Page looked at Brin. Brin looked at Page.
    A long silence. Finally: fine.
    They hired Eric Schmidt. Schmidt came from Novell — a textbook big-company manager.
    But his arrival brought an unexpected benefit — Schmidt saw the contradiction that Page and Brin hated advertising but had to make money.
    The three of them debated for weeks, then made a decision.
    No banner ads.
    Let advertisers write their own text ads and bid by click. The more relevant the ad and the higher its click-through rate, the higher its placement.
    Don't sell rankings to advertisers — sell positions to advertisers, but let the algorithm decide those positions based on relevance, not money.
    This was AdWords.
    Launched in October 2000. First year: zero revenue.
    Second year: $70 million.
    Third year: $1.5 billion.
    Google had become a money printer.
    But not because of advertising.
    Because they used algorithms instead of humans to decide what users would see.
    No editors. No reviewers. The code ran itself.
    This principle later permeated every corner of Google.
    Search results — ranked by algorithm.
    Ads — ranked by algorithm.
    Employee interviews — scored by algorithm.
    Page and Brin believed one thing: replacing human judgment with machines was not only cheaper, but more objective.
    That belief would be challenged again and again over the next twenty-plus years. But that's another story.

Now stop for a moment.
Reading this far, I want to ask you a question.
If you built a product today, and someone came along offering to buy you —
Would you sell?
You have no revenue. You can't pay for servers. Every day you survive on loose change from the vending machine.
Someone offers $750,000.
Would you take it?
99% of people would.
Page and Brin wanted to sell too. They just got lucky — they ran into a buyer unwilling to pay $750,000.
But what made Google Google wasn't that rejection.
It was what came after.
After being rejected, they didn't go looking for a second buyer.
They came back and rewrote the code.
They shaved another 0.2 seconds off the search speed.
That was their answer.

I've studied them for a long time.
Here's the plain truth. It comes down to three things again.
First: they didn't just build a search engine — they redefined what a search engine even meant.
Before Google, search = finding keywords.
After Google, search = finding authority.
This wasn't defined by editors. It wasn't defined by advertisers. It was defined by every website on the internet, casting its vote through links.
They didn't compete with Yahoo on who had the bigger directory. They made the directory category disappear altogether.
Is your current product improving its category, or redefining the category entirely?
Second: they baked an "unfair advantage" into the algorithm itself.
PageRank naturally favored older websites with many citations — the bigger the brand, the harder to unseat.
AdWords naturally favored advertisers who wrote better ads — placement depended on click-through rate, not bid amount.
When your business model inherently rewards "those who do it better," you no longer need to spend money maintaining rankings.
Does your product design contain this kind of positive feedback loop?
Third: they built an advertising company with "no middlemen."
Traditional advertising — client hires agency, agency buys media, layers of commissions, no one knows if the data is real.
AdWords — open a webpage, enter a credit card, ads go live. Results visible in real time. CTR, CPC, ROAS — all crystal clear.
They transformed advertising from a "relationship business" into a "data business."
In your industry, is there a "relationship barrier" just waiting to be dismantled by a "data barrier"?

That $750,000 rejection was the best thing that ever happened to Page and Brin.
It forced them onto a harder road —
Build the company themselves, write the algorithm themselves, set the rules themselves.
Twenty-six years later, the thing Excite rejected is worth $4.6 trillion.

← Previous Contents Next →
Chapter 3 Apple

Kicked Out of His Own Company for 11 Years, He Came Back and Did Something Even More Brutal: Cut 70% of the Product Line

Not innovation. Not R&D. Not marketing. Not distribution.
It was this: when the company was 90 days from bankruptcy, a man who had been cast out came back, and the first thing he did — cut.
Cut the printers. Cut the scanners. Cut the Newton PDA. Cut hundreds of SKUs.
Took 15 products down to just four.
One year later, Apple was alive.
Fourteen years later, Apple was the most valuable company on earth.
Let me tell his story first. After that, we'll talk about why most founders go their entire lives without ever learning how to "not do."

April 1, 1976, Los Altos.
In the Jobs family garage, three men hunched over a workbench soldering circuit boards.
Wozniak handled the design. Jobs handled yelling into the phone "not good enough." Wayne handled drafting the partnership agreement and drawing the first logo.
The machine was called the Apple I.
Selling price: $666.66.
First month, they sold 50 units. Made money.
But that's not the point.
The point is: 12 days later, Wayne walked out.
He sold his 10% stake for $800.
Today, that 10% is worth $430 billion.
Wayne later said in an interview:
"I don't regret it. I was 41 at the time. They were in their early twenties. I'd already been through one failed venture. I didn't want to lose money again."
You see, not everyone can bear uncertainty.
Some people's brains calculate probabilities. Other people's brains see — inevitability.

The Apple II launched in 1977.
The first true personal computer in human history.
Before this, computers were for engineers and corporations. Toggle switches. Punched paper tape.
The Apple II had a plastic case. Color graphics. A keyboard.
Within two years, Apple's revenue exploded from $775,000 to $117 million.
IPO in 1980. Jobs was 25, with a net worth of $217 million.
At that moment, he believed he was incapable of failure.
Then came 1984.

The Macintosh launched.
That famous Super Bowl ad — the woman in red smashing the Big Brother screen.
The whole world cheered.
The whole world didn't buy.
The Mac was too expensive. Starting price: $2,495 — roughly $7,400 in today's dollars.
Throughout the autumn of 1984, Apple's warehouses piled up with unsold Macs.
The financials bled red. The board grew nervous.
The power struggle between Jobs and then-CEO Sculley intensified, boiling over.
Jobs wanted to do whatever he wanted. Sculley wanted the company to have order.
The board met for an entire day.
The vote: 8 in favor of Sculley. Jobs was out.
The only vote cast for Jobs was his own.
He later spoke about that day at the Stanford commencement address:
"How could I, at 30, be fired by the person I brought in to run the company I founded?"
That night, he cried.
Then he did three things:
Sold all his Apple stock, keeping exactly one share — "to keep a lookout post."
Woke up at 5 a.m. the next morning and went running.
Founded a new company called NeXT.

The story of NeXT, in everyone's eyes at the time, was a joke.
Jobs built a black cube-shaped computer.
So perfect that every edge was molded as a single piece, made from magnesium metal.
Its operating system was called NeXTSTEP — which, as you now know, became the kernel of Mac OS X.
But no one knew that back then.
NeXT's computer sold for $6,500 — three times the price of an ordinary PC.
Nobody bought it.
The factory ran for 40 days before it was forced to shut down — zero orders.
The investors' money was burning fast.
At the same time, Pixar, the animation studio he had invested in, was bleeding cash too.
Year one: a $10 million loss.
Year two.
Year three.
Losses every single year.
He dug into his own pocket and poured in $50 million, keeping it afloat with his own money.
Anyone else would have given up by now.
Jobs's reaction — keep investing.
Not because he knew Pixar would hit big.
Because he believed in one very simple principle:
Technology + storytelling = something nobody can beat.
1995. Toy Story hit theaters.
Global box office: $373 million.
Pixar went public. Jobs's net worth multiplied tenfold overnight.
But NeXT still couldn't sell.
Then, December 1996.
Apple called.

Here's where Apple was at that moment:
Five consecutive losing quarters. Bleeding over $1 billion a year.
Marketing ran a survey asking consumers, "Would you still buy Apple products?"
The answer — "When your company goes bankrupt and has a liquidation sale."
CEO Gil Amelio made a decision.
He needed a new operating system, or Apple was dead.
He looked at two companies: one called Be, one called NeXT.
Be asked for $300 million. NeXT asked for $429 million.
Amelio chose NeXT.
Not because NeXT's system was better — because Jobs threw himself in for free.
January 1997. Jobs returned to Apple as an advisor.
Six months later, he ousted Amelio and made himself interim CEO.
He walked into Apple's headquarters lobby and saw a scene: 15 rows of product posters on the wall, and behind each row, dozens more folded under a "More" button.
Fifteen computer models. Dozens of peripherals. More printer models than many companies' entire product lines.
Not a single one you could remember.
He called the executives into the conference room.
On a whiteboard, he drew a four-quadrant grid.
Consumer desktop. Consumer laptop. Pro desktop. Pro laptop.
"We will make four products. Everything else stops."
Five seconds of silence in the room.
Then it exploded.
Someone from the printer division stood up — do you know how many printers we sold last year?
Jobs looked at him:
"That's exactly why we're cutting them."
That day, Apple cut 70% of its product line.
Three thousand people laid off.
Then he did something even more incomprehensible.
He flew to Seattle and walked into Bill Gates's office.
"I need $150 million."
Gates looked at him.
Jobs said: "Let's stop suing each other in court. You invest $150 million for Apple stock. We'll make Internet Explorer the default browser on Mac."
Gates agreed.
August 1997. Microsoft announced a $150 million investment in Apple.
At Macworld, Jobs put Bill Gates's giant face up on the screen behind him.
The audience screamed in anger. Boos thundered through the hall.
Jobs stood on stage. Didn't speak for a single second.
When the booing stopped, he opened his mouth:
"We need to let go of this."
"For Apple to get better."
A year later, the iMac launched. Colorful translucent shell. A handle. No floppy drive.
150,000 units sold on launch day.
Apple was alive again.

Now stop for a moment.
There's something here that 99% of founders will never do in their entire lives.
Not addition. Subtraction.
After Jobs returned to Apple, he didn't release a single new product.
Everything he did was cutting — cutting products, cutting projects, cutting people, cutting past grievances.
Not until the iMac launched in 1998 did he finally, for the first time, truly "add" something.
Do you have the courage to cut 70% of your company's business lines right now?
Delete 80% of the SKUs from your product list?
Do you have a list of "what not to do"?
If you only have a "to do" and no "not to do" — you're just accelerating the fragmentation of your own attention.

Jobs's story, at its core, is three things.
First: focus isn't doing one thing. It's not doing ten thousand things.
When Apple went from 15 products to 4, what got cut wasn't just products — it was entropy within the organization.
Every product carried a team, a VP, a pile of fixed costs. When nobody was buying those products, they were parasites sucking blood.
Jobs didn't pick the four best products. He killed everything that wasn't good enough.
How many SKUs do you have right now? How many business lines? How many "just in case it's useful" projects?
Second: what saves a company isn't a new idea. It's stitching up old wounds first.
He didn't do market research. No user interviews. No strategic committee.
He did one thing: stopped the company from bleeding.
Stopped making products that didn't make money.
Stopped fighting unwinnable lawsuits with Microsoft.
Stopped internal power struggles.
Before you "grow new flesh," have you stitched up the wounds that are still bleeding?
Third: those 11 years he was cast out — that wasn't loss. That was accumulation.
NeXT taught him how to build an uncompromising product.
Pixar taught him how to tell a story.
The humiliation of being cast out taught him how to run a company — this time, he wouldn't let his own CEO take him down.
Those 11 years were his moat.
Could the "low point" you're living through right now be building a fortress for you that no one else can see?

When Jobs died in 2011, Apple's market cap was $350 billion.
Today, $4.3 trillion.
What he left behind wasn't a company.
It was a way of thinking that taught the entire world to say "less is more."
That partner who walked out after 12 days, Wayne — he gave up 10%.
That 10% became the distance between him and being a billionaire.
But what truly made Apple Apple wasn't the 10% Wayne left behind.
It was the whiteboard Jobs returned to 11 years later.
On it, only four squares.

← Previous Contents Next →
Chapter 4 Microsoft

One Magazine Cover, Two Harvard Dropouts, and the Contract That Changed the World

Not inventing software. Not writing code. Not building the first operating system.
It was this: when they had no product at all, telling IBM "we have one."
Then spending $50,000 to buy a system called QDOS and renaming it MS-DOS.
The crucial move — convincing IBM to let Microsoft keep the right to sell DOS to every other PC maker.
With that single move, he created one of the most valuable business clauses in human history.
Let me tell his story first. After that, we'll talk about what true business insight really is.

December 1974, Boston. A blizzard.
Paul Allen burst into a Harvard dorm room clutching a magazine.
Popular Electronics, January 1975 issue.
On the cover: a miniature computer called the Altair 8800. A rectangular box, its front panel lined with tiny lights and toggle switches.
The world's first "personal computer."
Allen slapped the magazine onto Bill Gates's desk.
"Look. The thing we've been talking about — someone's built it. And we — we haven't written a single line of code."
Gates glanced at the cover.
He picked up the phone and dialed the number of MITS, the Altair's manufacturer.
"Hello. We've developed a BASIC programming language for the Altair. Are you interested?"
They said: absolutely. Send it over and we'll take a look.
There was just one problem — they hadn't written a single line of code.
Gates hung up and looked at Allen.
"We have two months. Let's start."

The two of them launched into a death-defying marathon.
Gates wrote the core code. Allen wrote the simulator — because there was no real machine to test on.
They slept in shifts. One wrote at the keyboard until he collapsed. The other took over.
Harvard CS students walking past Gates's dorm would hear the keyboard still clattering at three or four in the morning.
Someone pulled back the curtain to look — Gates was slumped over the keyboard, already asleep. His fingers still resting on the ASDF keys.
Eight weeks later, Allen flew to Albuquerque, New Mexico, carrying a roll of punched paper tape. He walked into MITS.
A row of engineers sat before him.
He fed the tape into the reader.
Seconds later, a single word appeared on the terminal:
"READY."
For the first time, the Altair could be programmed in BASIC.
March 1975.
Gates was 19 that year.
That autumn, he dropped out of Harvard, flew to Albuquerque, and with Allen founded "Micro-Soft" — the hyphen was later removed.
All right.
That's just the beginning.
What truly turned Microsoft into a giant was the phone call that came in the summer of 1980.

IBM was working on a secret project, code-named "Chess."
They were going to build a personal computer — the IBM PC.
They needed an operating system.
The IBM team first went to a company called Digital Research, run by a man named Gary Kildall.
Kildall's CP/M operating system was the industry standard at the time.
The IBM representatives showed up at Kildall's front door and knocked.
Kildall wasn't home. He had gone flying his plane.
The IBM people turned away, faces grim.
Then they came to Microsoft.
But Microsoft didn't make operating systems. Microsoft made programming languages.
Gates did not say "we don't do that."
He said — "We can."
He made a phone call to Seattle Computer Products.
For $50,000, he bought a system called QDOS.
QDOS stood for — "Quick and Dirty Operating System."
It was basically a knockoff of CP/M.
Gates renamed QDOS to MS-DOS, and then walked into IBM's boardroom.
At the signing, Gates insisted on a single clause in the contract:
Microsoft retained the right to license MS-DOS to other computer manufacturers.
IBM's legal counsel glanced at it and shrugged. "Sure. The PC market won't be that big anyway."
That sentence cost IBM roughly $2 trillion in future value.
Because here's what happened next:
Compaq, Dell, HP, Lenovo — hundreds of manufacturers flooded into the PC market.
Every one of them built their own machines. Every one of them needed an operating system.
And there was only one company that could give them one — Microsoft.
From 1981 to 1990, over 100 million PCs were sold worldwide.
More than 80% ran MS-DOS, and later, Windows.
Microsoft wasn't a company that sold software.
Microsoft was a company that collected a "PC tax."
Every PC user, no matter what brand of machine they bought, ultimately paid Bill Gates.
This is the most exquisite move in business history — you're not selling a product. You're defining who collects the toll.

But there's one thing here that everyone overlooks.
Gates wasn't lucky. IBM wasn't stupid.
In 1976 — four years before that call from IBM — Gates wrote an open letter.
Its title: "An Open Letter to Hobbyists."
In it, he tore into everyone who copied Microsoft's software for free.
The core line: "If software developers don't get paid, who will write software?"
In 1976, this idea was heresy.
Every coder at the time believed software should be free. Hardware had value. Code? What was code?
But Gates saw something that practically no one else saw —
Software could be commoditized.
Code had copyright.
A line of text, once written, could be copied infinitely. The marginal cost of every additional copy approached zero.
No one in the world understood this business model earlier than Gates.
He was 19 years old, and he already saw it clearly.

Now stop for a moment.
There's a question worth pausing to think about:
Could you, at age 19, see a business model with perfect clarity — and then bet your entire life on it?

I've studied Gates's story. Here's the plain truth. Three things.
First: it wasn't product first. It was order first.
Gates called MITS before he had any product.
He promised IBM an operating system before he had any operating system.
What he did consistently was — lock down the buyer first, then find the supplier.
This is the highest form of "middleman" thinking.
But what do most founders do?
Build the product first, then go looking for someone to buy it.
Spend six months building. Spend another six months hunting for customers.
Gates did the reverse. Order in hand, then go find the supply.
In your current development process, do you lock in demand first, or build the product first?
Second: he understood the role of the "component supplier."
In the auto industry, no one can monopolize the whole car because parts come from so many different suppliers.
In the PC industry, Intel monopolized the CPU. Microsoft monopolized the operating system.
Why?
Because the CPU and the operating system are the "standard components" that every PC needs.
Whoever controls the standard components controls the entire industry.
He never positioned himself as "the best software company."
He positioned himself as "the only operating system supplier in the PC industry."
In your industry, is there a "component" that everyone needs but no one has monopolized yet?
Third: copyright — how do you use it to create real value?
That open letter in 1976 — everyone attacked him for it at the time.
But Gates saw one thing clearly — code is intellectual property, and intellectual property can be licensed.
A software company doesn't sell products. It licenses usage.
That one-word difference — "sell" versus "license" — created the entire business model of the software industry.
In your current business model, is there a "sell" that could become a "license"?

Bill Gates probably still looks at the digital edition of that Popular Electronics magazine this year.
Fifty-one years ago, it was the photograph on that magazine cover that made two young men decide to drop out and write software that didn't yet exist.
They weren't building a product.
They were inventing an industry.

← Previous Contents Next →
Chapter 5 Amazon

Nine Years of Losses, Stock Crashed from $113 to $6 — and He Did Something Even Crazier: Kept Investing

Not optimizing profits. Not tightening the belt. Not appeasing Wall Street. Not laying off to survive the winter.
It was this: every morning when he opened the newspaper to headlines screaming "Amazon is doomed," one man, holding his parents' entire life savings, kept pouring money into warehouses.
He poured for nine years.
And then he built his company into a $2.8 trillion empire.
Let me tell his story first. After that, we'll talk about why "losing money" is sometimes the smartest strategy there is.

Summer 1994, New York, Manhattan.
D.E. Shaw hedge fund, 40th floor.
Jeff Bezos sat at his desk, staring at a single number on his screen.
Internet users, annual growth rate: 2,300%.
He later recalled the feeling of that moment in an interview:
"I saw something growing at 2,300% a year. I had never seen anything grow that fast."
"I knew that if I didn't jump in, I would regret it for the rest of my life."
He built a "Regret Minimization Framework" — sounds sophisticated. In plain English:
When I'm 80, I won't regret trying and failing.
I will regret never having tried at all.
The next day, he quit.
His boss, D.E. Shaw founder David Shaw, said to him:
"Jeff, you're a very good hedge fund analyst. Are you sure you want to leave?"
Bezos said: I'm going to sell books on the internet.
David Shaw paused.
"That idea sounds... doable. But I think it's a better fit for someone who doesn't already have a good job."
Bezos smiled.
The next day, he and his wife MacKenzie threw their luggage into the trunk of a 1988 Chevy Blazer.
They set out from New York, heading west.
MacKenzie drove. Bezos sat in the passenger seat, typing a business plan on his laptop.
They stopped in Texas to see his parents. He showed them the plan.
His parents were quiet for a moment.
Then they wrote a check for their entire life savings — $300,000 — to their son.
"We're not investing in Amazon. We're investing in you."
July 1994. They rented a house in Seattle.
The garage was the office.
Bezos personally built two desks out of door panels.
That was the beginning of Amazon. Two door panels and $300,000.
From 20 product categories, he narrowed it down to books.
Why?
Not because he loved books.
Because books don't rot. Their prices are easy to scan. Shipping is cheap. And there are 3 million books in print worldwide — the largest physical bookstore could hold at most 100,000.
The internet could hold all 3 million.
This was something the physical world simply could not do.
The first month, they sold books to all 50 U.S. states and 46 countries.
Only a few boxes of books sat in the warehouse, but the website listed one million titles for sale.
Customers placed orders. They went and found the books.
May 1997. Amazon went public. Stock price: $18.
Bezos wrote his first shareholder letter.
One line from that letter lodged itself in everyone's memory: "Today is Day 1."
And then, every year, he wrote another shareholder letter, and at the end of each one, he attached that 1997 letter.
To this day, he has attached it 29 times.

Then came the year 2000.
The dot-com bubble burst.
Amazon's stock plummeted from $113 all the way down to $6.
Wall Street analysts ran daily headlines — "Amazon: The Next Dot-Com to Go Bankrupt."
Barron's cover story was titled "Amazon.bomb."
In Bezos's office, there was a whiteboard.
The day the bubble burst, he didn't write a single number on it.
He drew a flywheel.
He wasn't calculating how to survive the quarter.
He was calculating: if we grow the user base → we get lower wholesale prices → lower prices attract more users → more users attract more third-party sellers → more sellers expand selection → richer selection draws more users.
Once this loop started spinning, it wouldn't need to be pushed every day.
At a time when everyone was screaming "cut costs," he did three things that everyone thought were insane:
First, lower prices. Not raise them. Lower them. Slash prices across the board.
Second, keep building warehouses. Between 2000 and 2002, he built over a dozen new fulfillment centers.
Third, bring in third-party sellers. Expose his own customers to his competitors.
The board suggested he at least raise prices a little — make the profits look a bit better, get the stock off that $6 floor.
Bezos said: no.
"Profit margin is not our goal. Free cash flow is."
"Every penny saved today, if not reinvested in the customer experience, is stealing from tomorrow's growth."
2003. Amazon posted its first full-year profit.
Exactly nine years had passed since he quit his job to sell books.
But the real plot twist came in 2006.

That year, Bezos did something that left everyone utterly baffled.
Not because it was related to e-commerce. Because it had nothing to do with e-commerce at all.
He launched Amazon Web Services — cloud computing.
Rent out the company's spare server capacity and storage — let other people run their websites and applications on it.
Why would a bookseller do this?
Investors began to question: has he lost focus?
Bezos ignored them.
First, they rented to startups — small teams without the money to build data centers.
Then mid-sized companies started arriving.
Then Netflix came and moved all its servers onto AWS.
Then NASA came.
Then the CIA came.
AWS started in 2006. By 2024, its annual revenue exceeded $90 billion.
To this day, AWS contributes over 60% of Amazon's operating profit.
And this business's only connection to "selling books on the internet" is — they both need servers.
But Bezos saw something in 2006:
Internet companies don't need to buy their own servers.
Just like factories don't need to build their own power plants — you just plug into the grid.
AWS is the internet's "computing grid."
He spent a full seven years waiting for the world to accept this logic.

Now stop for a moment.
There's something I want you to pause and think about carefully.
When everyone was yelling at you to "control costs," "turn a profit," "stop burning cash" —
How many times did you do what they said?
Have you ever experienced a moment — you knew you were right, but every external signal was telling you to stop, and you chose to keep going?
If you listened every time —
Maybe you've been missing your AWS.

I've studied Bezos for a long time. Three things:
First: he looks at variables 20 years out, not the stock price 20 days out.
Bezos has a line that gets quoted endlessly: "I'm often asked: what will change in the next 10 years? But almost no one ever asks me: what will not change in the next 10 years?"
His answer: customers will always want lower prices, faster delivery, and more selection.
So he poured all his money into those three things — price, logistics, selection.
Will logistics change 20 years from now? No. So he sank over $100 billion into logistics.
In your industry, what will not change in 20 years?
Second: it's not "diversification." It's multi-layered infrastructure.
E-commerce is the land. Third-party sellers are the storefronts. AWS is the power grid.
Every business is infrastructure at a different layer. Once built, others have to pay to use it.
Is your business selling products, or building infrastructure?
Third: "regret minimization" isn't fortune-cookie wisdom. It's a decision-making tool.
The standard for whether to do something isn't "will it succeed?" — it's "will I regret not trying?"
Stretch the timeline to your 80th year. Which choices will you regret never having made?
Those are the ones you should be making today.

Bezos probably still keeps that desk made from a door panel.
Every time new employees start, HR takes them to see that garage.
The door panel is very old now.
But the words written on it — Day 1 — have never been crossed out.

← Previous Contents Next →
Chapter 6 Broadcom

He Acquired 40 Companies — Not One of Them His Own — and Built a $2 Trillion Chip Empire

Not starting from a garage. Not writing code. Not inventing new technology.
It was a Chinese-Malaysian immigrant, fresh out of MIT, using a method no business school teaches — buy.
Buy 40 undervalued companies. Slash every business line that doesn't make money. Squeeze maximum cash flow out of every single one.
Thirty years later, Broadcom, under his helm, has a market cap of $1.97 trillion — sixth-largest in the world.
Let me tell his story first. After that, we'll talk about why subtraction is a hundred times harder than addition.

His name is Hock Tan.
Born in 1952 in Penang, Malaysia, to a Chinese family.
They weren't wealthy. But he was exceptionally gifted at math.
At 18, he won a scholarship to MIT. Studied mechanical engineering.
Later, he earned an MBA from Harvard.
After graduation, he went to General Motors, then PepsiCo, then Commodore International.
His first job was at GM — managing budgets, not managing technology.
That decision shaped his entire mindset for life.
He watched Detroit's auto giants burn tens of billions of dollars every year developing new cars, while their profit margins were paper-thin.
Meanwhile, over at Coca-Cola — a company selling sugar water — gross margins were 60%.
A question began to crystallize in his mind:
"Why do some companies fail to make money — not because they're doing too little, but because they're doing too much?"

In 2006, he joined Avago Technologies — a chip company spun off from Hewlett-Packard.
Then he set in motion the most ferocious M&A machine in history.
2013: bought LSI Logic. Immediately after the purchase, cut the money-losing divisions. Focused on storage chips. Profits doubled.
2015: bought Emulex — a networking chip company. Laid off 20%. Focused on core customers. Profits tripled.
2015: bought his former competitor — Broadcom. The day the acquisition closed, he renamed the company Broadcom. Not because he wanted the brand — because that brand was worth more than Avago. But he also reshaped Broadcom's style into his own.
2017: he attempted to acquire Qualcomm for $130 billion. Had it succeeded, it would have been the largest M&A deal in tech history. The Trump administration blocked it on "national security" grounds.
2018: he bought CA Technologies for $18.9 billion — a legacy company that made mainframe software. Wall Street went insane. Everyone asked the same question: why would a chip company buy a firm that builds 1960s mainframe software?
Hock Tan didn't answer.
He flew to CA's headquarters and summoned every salesperson into the conference room.
"How many of you have customer renewal rates above 90%?"
Nearly every hand went up.
"Starting tomorrow, all R&D projects unrelated to those customers — shut them down."
"You are not a technology company. You are a — "
He paused —
"Cash cow."
The R&D budget was slashed by 60%. Profit margins surged from 20% to over 40%.
CA became Broadcom's most stable source of cash.
2019: bought Symantec's enterprise security business — $10.7 billion.
2022: pulled off another of the largest deals in tech history — acquiring VMware for $61 billion.
Same formula: buy it, cut the peripheral businesses, focus on the core customers, raise prices.
Under his governance, VMware's operating margin leaped from 30% to over 70%.
Same recipe, every single time.
After every acquisition, Wall Street said he was insane.
Then after every acquisition, the numbers he delivered looked better than the previous quarter.
He never invented a single technology.
He never created a single product.
He created an entirely new method of corporate governance —
Don't chase the technological frontier. Chase the perpetual motion of cash flow.

Now stop for a moment.
Have you ever thought about this question?
In your business, is there a product line you already know is unprofitable — but you're still feeding it?
Not because of strategy. Not because it's going to grow. Because — you built it yourself, and cutting it breaks your heart.
Hock Tan used a lifetime of experience to tell you this: the things you can't bear to cut will eventually cut you down.

Hock Tan's methodology is simple to describe, but vanishingly few people can execute it:
First: he doesn't buy technology. He buys "locked-in customers."
CA, VMware, Symantec — these companies share one trait.
Their software is bolted into the core systems of the world's biggest corporations. Decades of accumulated code. The cost of switching is astronomical.
A bank's core system has been running on CA's software for 30 years.
Want to switch? Go ahead. You'll need five years, a billion dollars, and you'll risk crashing every production system in the process.
So the customers don't switch.
So Hock Tan can raise prices — not by 5%. By 50%. By 200%.
And you just have to take it.
Your current customers — what is their cost of leaving you?
Second: he doesn't pursue innovation. He pursues "profit without innovation."
This runs completely counter to Silicon Valley's DNA. Silicon Valley's creed is — innovate constantly, or die.
Hock Tan's creed is — find a customer base that can thrive even without innovation, and then raise prices.
He once said something brutally honest:
"Innovation carries risk. Raising prices — carries no risk."
Is there a part of your business whose value hasn't yet been unlocked by the simple act of "raising prices"?
Third: he turned M&A into an assembly line.
Hock Tan doesn't buy one company and stop. He buys every year.
His targets share three characteristics: entrenched market position, extremely high product stickiness, and management teams that are terrible at cost control.
After the purchase, replace the management, cut costs to the bone, and turn the savings into cash flow.
He spent 30 years buying his way to a $2 trillion empire.

Hock Tan is 74 now.
He's still buying.
Go look at Broadcom's acquisition history — not a single year is blank.
He's not running a chip company.
He's running a business model where "someone always pays, and nobody can ever leave."

← Previous Contents Next →
Chapter 7 Tesla

Christmas Eve 2008 — He Had Only $200,000 Left, and He Burned It All Into Tesla

Not a technological breakthrough. Not government subsidies. Not market timing. Not luck.
It was this: at a time when every automaker was laughing at "electric cars are a joke," a serial entrepreneur from South Africa bet his entire net worth — hundreds of millions earned across four startups — on a company selling only 2,000 cars a year.
Then he built Tesla into the most valuable automaker on earth, its market cap at one point exceeding the next ten car companies combined.
Let me tell his story first. After that, we'll talk about why "going all in" is the rarest quality a founder can possess.

December 2008, Silicon Valley.
Elon Musk did the math.
Two companies under his name were about to miss payroll — Tesla and SpaceX.
What was left in his pocket, if split between them — both would die.
All to SpaceX, Tesla dies.
All to Tesla, SpaceX dies.
His net worth at the time — mostly paper wealth — had been virtually wiped out. The remaining cash: roughly $2 million.
He called Tesla's CFO into his office.
"Divide my money between the two companies. All of it. Every last dollar."
The CFO looked at him: are you serious?
Musk didn't answer.
The CFO said: that means if your next funding round doesn't come through... you will be personally bankrupt. Come Christmas, you won't even afford a plane ticket home.
Musk said: "Then bankrupt it is."
Christmas Eve. He sat alone in SpaceX's Los Angeles factory. A cup of coffee in his hands. Outside — no Christmas tree, no lights.
He later wrote one line in his biography:
"If you had asked me in 2008 what the probability of success was — I would have said 10%."
"But 10% is enough."

Stop here for a moment. I need to make clear what it is he was actually doing.

  1. Musk put $6.5 million into Tesla — an electric vehicle startup with five employees.
    Electric cars before this — golf carts. Top speed of 25 miles per hour. Nobody considered them actual "cars."
    Musk's bet: build a car that looked like a sports car, drove like a sports car, and ran purely on batteries — no gasoline.
    The auto industry called him naive.
    Detroit's engineers said lithium batteries could never deliver enough range.
    The host of Top Gear called Tesla a scam.
    His friends sent emails begging him not to be an idiot.
    He didn't reply.
  2. The Roadster finally entered production.
    But the manufacturing cost per vehicle was $120,000. The selling price: $109,000.
    Every car sold lost $11,000.
    And that year, he sold only 2,000 of them.
    The money on the books would last only a few more weeks.
    He went to see Daimler.
    After the meeting, someone from Daimler said one thing:
    "Your people talked about the technology for three hours. We didn't understand a word of it. But we understood what we saw in your CEO's eyes."
    Daimler invested $50 million.
    That was the last straw — the one that didn't break.
    Tesla didn't die in 2008.
  3. Tesla went public. The first American car company to IPO on the U.S. stock market since Ford in 1956.
  4. Model S launched. Consumer Reports gave it the highest score in history: 99 out of 100.
  5. Model 3 became the world's best-selling electric vehicle.
  6. Tesla's market cap surpassed Toyota — a company selling 10 million cars a year. Tesla had sold fewer than 500,000.
  7. Market cap crossed $1 trillion.
  8. Over $1.6 trillion.

But flip it over. I need to tell you the other side of those numbers.
From 2008 to 2018, Musk went through at least five "Tesla is about to go bankrupt" moments.
Not a metaphor. Real moments — two weeks from missing payroll.
Each time, he started from zero, knocking on doors for funding.
Each time, every investor said yes at first. Then no at the end.
Each time, he plugged the hole with his own credit.
He personally borrowed millions from friends just to pay his own rent.
August 7, 2018. He tweeted about "considering taking Tesla private at $420." That tweet got him fined $20 million and forced him to step down as Tesla's chairman.
He said in an interview:
"This past year has been the most painful year of my career. So painful. I slept on the factory floor."
"Not to show off. I genuinely didn't have time to go home."
"My kids looked at me — Dad, why don't you come home?"

Now stop for a moment.
I want to ask you: if it was 2008 and it was you — two companies about to collapse, only your last pile of cash left in your pocket. Would you burn all of it?
Have you ever gone all in on a single direction — bet everything you had?
Not most of it. Not 50%. Everything.
If you've never had that moment —
Then you may have never truly believed in anything.

Three things.
First: he wasn't gambling. He was doing a calculation that no one else was willing to do.
When everyone said there was no market for electric cars, what Musk saw was — the physical limits of fossil energy.
A gasoline engine's maximum thermal efficiency is 40%. An electric motor's is 90%.
This isn't a subjective judgment. This is the law of thermodynamics.
The laws of physics stand on the side of electric vehicles. Everything else is just a question of time.
So he wasn't gambling. He was calculating — when efficiency differs by an order of magnitude, no technological competition is fair.
In your industry right now, is there a "law of physics"-level variable that everyone is ignoring?
Second: he burns his escape routes every single time.
If Tesla burned, he'd be broke. If SpaceX burned, he'd be broke too.
That was by design.
He never keeps a Plan B.
Why would you? Keeping one is a temptation — when Plan A hits trouble, you'll run back to Plan B.
Burn the escape route. You have only one direction left. Then you'll fight like hell to make it work.
Do you have a Plan B? If you canceled it, would you fight harder on Plan A?
Third: his "pain tolerance" is ten times that of any competitor.
Every automaker CEO looks at Tesla's financials and says "this man is insane."
But Musk is not competing with GM or Ford. He's competing with time.
Whoever can endure losses the longest will be in the right place at the right time.
Tesla lost money for 17 years before reaching stable profitability. Which traditional automaker could endure 17 years of losses?
Is your pain threshold high enough to persist in a single direction longer than everyone else?

The hand Musk was dealt was never good. He just went all-in every single time.
Not because he has a gambler's instinct.
Because he had done the math — the upside on this bet was too high. Betting small would be the biggest loss of all.
And how big you can bet depends on how deeply you believe.

← Previous Contents Next →
Chapter 8 Meta

One Night at 19, He Built a Website in His Dorm. 21 Years Later, 2.7 Billion People Use It

Not a business plan. Not a technological revolution. Not user research. Not financial engineering.
It was a Harvard sophomore, freshly dumped by his girlfriend, who drank a few beers, sat in front of his dorm computer, and spent one night writing a few lines of code.
The website was called FaceMash — it put photos of Harvard women side by side and let users vote on who was hotter.
That same day, Harvard's campus network crashed. Because in four hours, 22,000 people visited that site.
Two years later, he dropped out of Harvard and renamed the thing Facebook.
Twenty-one years later, its market cap is $1.5 trillion.
Let me tell his story first. After that, we'll talk about why "boring" makes the best product.

A Wednesday night, October 2003. Kirkland House, room H33.
Mark Zuckerberg sat at his desk. Freshly broken up. In a terrible mood.
He drank a few beers. Opened his laptop.
First, he hacked into Harvard's dorm housing databases and scraped the student photos.
Then he wrote a simple PHP script — two photos side by side, one click for "hotter."
FaceMash went live at 10:42 p.m.
By 2 a.m., Harvard's campus network had buckled under the traffic.
The school's IT department was scrambled for an emergency investigation.
The next morning, the Harvard Crimson's front page headline: "Zuckerberg Creates Pornographic Site Attacking Women."
Harvard's Administrative Board launched an investigation to expel him.
At the hearing, Zuckerberg said one thing:
"I just thought the idea was interesting. I didn't know people would like it this much."
He wasn't expelled — but came terrifyingly close.
This was later made into the movie The Social Network.
But the movie only captured half the story.
What truly made Facebook Facebook was what came next.

After FaceMash was shut down, Zuckerberg's inbox filled with emails, one after another.
From Harvard students. Not here to curse him out — here to ask him: could you build something similar, but instead of comparing who's hotter, something that lets people... actually get to know each other?
Every semester, 1,600 new students arrived. No one could remember all the names.
You'd see someone in the dining hall — "Wait, what's his name?" "Can I sit with him for lunch?"
Zuckerberg listened.
This time, he wrote real code. Users had profiles, profile photos, course lists, friends lists.
He called it TheFacebook.
Launched February 4, 2004.
Within 24 hours, 1,200 Harvard students signed up.
Within 48 hours, it spread to every dorm at Harvard.
Within a week, students at Stanford, Yale, and Columbia were sending emails: "Open it up for us too."
A month later, TheFacebook covered more than 30 schools.
At this point, Zuckerberg faced his first counterintuitive choice:
Turn Facebook into a moneymaking machine — insert ads, charge membership fees — or keep it pure, and just build connections between people?
Every Silicon Valley investor gave him the same advice: make money, now.
Zuckerberg said no.
Why?
Because he saw something — this wasn't a website. This was a social graph.
How are real people connected to each other in the real world? Facebook would be the digital version of that relationship map.
Once this graph was complete, every way of making money would be a downstream byproduct of it.
What he saw was: the social graph must be complete first, and only then could commercialization become possible. Not the other way around.
He refused to put anything on Facebook that would interfere with the construction of that graph.
In 2004, the Facebook homepage bore a handwritten tagline from Zuckerberg:
"A Mark Zuckerberg Production."
In 2005, it changed to: "Facebook is a social utility that connects you with the people around you."
That wasn't marketing copy.
That was his definition of the product's essence — a utility.

But the real turning point came in September 2006.
Facebook did something every Silicon Valley veteran said was insane.
News Feed launched.
Before this, Facebook was basically a digital phonebook. You had to click on each friend's profile, one by one, to see their updates.
News Feed reversed the process — you didn't have to go looking. Facebook automatically arranged all your friends' activity into a stream and pushed it directly in front of you.
Launch day.
Users exploded.
Hundreds of thousands flooded into a protest group titled "News Feed Is the Most Privacy-Invasive Feature in History."
The New York Times, Washington Post, CNN — all negative coverage.
Protesters stood outside Zuckerberg's office holding signs: "Mark Zuckerberg Is a Creepy Stalker."
The board held an emergency meeting. Investors advised shutting down News Feed immediately.
Zuckerberg said: no.
He said something that would be endlessly quoted as a product philosophy:
"What users are afraid of isn't the feature itself. It's that they're not yet used to the idea that information can flow this way."
"But once they get used to it, there's no going back."
He didn't shut down News Feed. He added a privacy settings entry point.
Two weeks later, the protests faded.
Facebook's user engagement doubled.
From that day forward, Facebook went from "digital phonebook" to "something you check 15 times a day."
This is the core — he wasn't building a "better address book." He was building a product that made information-stream addiction.
Later, Twitter, Instagram, TikTok — all did the same thing.
He could see what others couldn't.

2004 to 2012. Facebook took zero money. Kept building connections.
2012. IPO. Market cap broke $100 billion.
Same year, bought Instagram for $1 billion — a photo-sharing app with only 13 employees. Wall Street said he was insane.
2014. Bought WhatsApp for $19 billion. Wall Street said he was even more insane.
2021. The company renamed itself Meta. Went all-in on the metaverse. The stock dropped 70% that year. Wall Street said he had completely lost his mind.
But Zuckerberg's response — after falling 70%, he kept investing in 2023.
The board asked: can we at least reduce the metaverse investment a little?
He said: no.
By the end of 2023, Meta's stock had climbed from $88 to over $400.
Wall Street shut up.

Now stop for a moment.
There's something you should notice.
Every time Zuckerberg went against the market, it wasn't because he didn't know what the market was thinking.
It was because he was looking at the communication infrastructure of the next decade, and the market was looking at the financial statements of the next quarter.
What's the most recent decision you made that went "against the market"?
If there is none — have you just been listening to others telling you which way to walk?

Three things.
First: he wasn't building a social network. He was building a social graph.
A social network can be copied. MySpace was copied to death. Google+ was copied to death.
But a social graph — the real relationships between you and everyone you know — no one can copy that. It can only be accumulated.
You have 1,000 friends on Facebook. Ten years of photos. Two hundred groups. You can't leave.
Not because Facebook's product is better. Because your social relationships exist on that one server, and nowhere else in the world.
Does your product record data that makes it impossible for users to leave?
Second: he turned "boring" into the most powerful user habit of all.
The essence of News Feed isn't that the content is amazing — it's that you can scroll mindlessly.
Like eating potato chips. No single chip is amazing, but you can eat the whole bag.
Boring = low barrier to entry = high frequency = habit = moat.
Have you ever made a feature so boringly simple that users can repeat it without thinking?
Third: he didn't spend $19 billion buying companies. He was buying the things that could one day disrupt him.
Instagram. WhatsApp. Oculus. At the time, every one was mocked as "an insane price for something that doesn't make money."
But Zuckerberg bought them not because they were valuable right now. His judgment was — ten years from now, the way people communicate will be completely different.
He wasn't buying products. He was buying directions.
What have you bought based not on current cash flow, but on where things are heading ten years from now?

Zuckerberg is 42 now.
The website he accidentally built at 19 — 2.7 billion people open it every day now.
Not because the code was well-written.
Because every time the whole world opposed him, he chose to believe what he alone could see.

← Previous Contents Next →
Chapter 9 Berkshire-Hathaway

60 Years, $1 Trillion — He Did Only One Thing Slower: Nothing

Not frequent trading. Not technical analysis. Not diversified hedging. Not leverage.
It was this: sitting in an office in Omaha, reading 500 pages of financial reports every day, making just a handful of investment decisions each year, and then — waiting.
He wasn't trading stocks. He was waiting for time to do the work for him. He waited for 60 years.
And he became one of the wealthiest people on earth. His company's market cap exceeded $1 trillion.
Let me tell his story first. After that, we'll talk about why "holding still" is a hundred times harder than "constantly moving."

1956, Omaha.
Warren Buffett was 26.
He had graduated from Columbia Business School and studied under Benjamin Graham for two years. Then he returned to his hometown of Omaha and founded a partnership investment firm in a small upstairs room of his house.
Starting capital: $105,000. One hundred dollars of it was his own.
His first client was his Aunt Alice, who put in $25,000. Then his father-in-law, his college roommate, his childhood neighbor.
He told all of them one thing:
"I'm not here to trade stocks for you. I'm here to buy companies for you."
"If you care about day-to-day ups and downs — don't invest with me."
"If you're willing to let me prove myself over five years — then sign."
Everyone signed.
Then he did something that confused everyone.

He started buying a textile mill called Berkshire Hathaway.
This was 1962. America's textile industry was being crushed by Japan.
Every textile mill was closing. Berkshire was no exception — based in New England, aging equipment, hemorrhaging workers. Losing money every year.
Buffett's partners called to ask: "Why are you buying a dying textile mill?"
Buffett didn't explain much.
His logic was simple — this company's net assets, after depreciation, were worth far more than its stock price.
He wasn't buying a business. He was buying something cheap.
He kept buying until he had a controlling stake. Then he replaced the management and tried to turn things around.
The result?
He failed.
The textile mill kept losing money. The Japanese were too strong. American labor costs were too high. No matter what he tried, he couldn't flip it. Shutting it down was also too expensive — severance, pensions, factory demolition, a fortune.
Buffett turned Berkshire into his biggest mistake.
He later wrote in his annual report: "I should not have bought Berkshire. It was an impulsive mistake."
But it was this very mistake that became the most important inflection point of his life.
In 1985, he finally shut down the textile operation.
But he kept the name Berkshire and its legal structure.
He took all the cash recovered from textiles and started buying in a new direction — insurance companies.
Why?
Because insurers have something called "float." The premiums you prepay when you buy insurance — until a claim is made, the insurance company holds that money.
If you can invest that money before it has to be paid out in claims — you're using other people's money to make money, for free.
The first insurance company Buffett bought was called National Indemnity. 1967.
Then GEICO — he'd bought GEICO stock when he was 20, then sold it. Now he bought it back, for $2.3 billion.
Then General Re — one of the world's largest reinsurers. 1998, for $22 billion.
Float grew from $20 million in 1967 to over $160 billion today.
He now had a perpetual cash machine.
And with that machine, he made a series of moves:
1988: bought Coca-Cola. Spent $1 billion. Today, that investment is worth $23 billion. It pays over $700 million in dividends every year.
1989: bought Gillette.
1990: bought Wells Fargo.
2008 financial crisis: when everyone was selling, he single-handedly invested $5 billion in Goldman Sachs. It nearly doubled in five years.
2016: he did something everyone thought meant "he's gotten old" — bought Apple. By then, Apple was already the world's biggest company, and everyone was saying Apple's best days were behind it.
By 2025, that Apple investment is worth $170 billion.
But his most famous trade was actually a "rejection."

Year 2000. Peak of the dot-com bubble.
Every fund manager was buying dot-com stocks. Every single one.
Buffett's partners told him: "You have to buy. If you don't buy tech stocks, you're going to be left behind."
His shareholders wrote letters cursing him, saying he was obsolete. The media ran cover stories: "Has Buffett Lost His Touch?"
Buffett read all of it. And did one thing.
He didn't buy.
He kept buying American Express. Coca-Cola. Wells Fargo. Companies that had already existed for a hundred years.
He later said something very famous:
"I don't invest in things I don't understand — that's not conservatism. That's knowing the boundary of your own competence."
Then, in 2001, the bubble burst.
The Nasdaq crashed from 5,048 to 1,114. Trillions of dollars in global tech stocks evaporated overnight.
Buffett was untouched.
By "not buying," he had gained an entire bull-and-bear cycle's worth of time over everyone else.

By 2019, Berkshire Hathaway was sitting on over $128 billion in cash. Buying nothing.
A CNBC anchor asked him: why aren't you investing? What are you waiting for?
Buffett said: "I'm waiting for an elephant."
An "elephant" meaning — an investment opportunity good enough.
Then came March 2020. COVID hit. Global stock markets fell 35% in two weeks.
While every CEO was in emergency meetings figuring out how to survive, Buffett opened his phone book and started buying.
He had waited many years for an elephant — and the elephant walked right into his living room.

Now stop for a moment.
Do you have a sum of money that's been sitting in your account since last year, uninvested — simply because you're waiting for an opportunity big enough?
If you always deploy every dollar you have, you'll never catch the elephant.
Because he's not waiting for a "decent" opportunity. He's waiting for a "once in a lifetime" price.
Can you, like him, do nothing — until the elephant appears?

Three things.
First: he's not an investor. He's a "capital allocator."
An investor asks every day: "When should I buy? When should I sell?"
A capital allocator asks: "How much capital do I have? How do I allocate it to the places with the highest returns over the next 10 years?"
One looks at price movements. The other looks at capital allocation efficiency.
Which one do you look at?
Second: he doesn't hunt for opportunities. He waits for opportunities to come to him.
Berkshire always has tens of billions in cash on its books. That's not waste — that's ammunition.
When a plague, a crisis, a panic hits, everyone else is on the phone scrambling to raise money. Buffett doesn't need to make a single call. He just takes a slice from $128 billion.
Can you maintain an "arsenal" — a large enough cash reserve — reserved solely for waiting for those rare, extreme opportunities?
Third: his circle of competence never expanded — but his wealth kept compounding.
He never invested in the internet (until buying Apple in 2016 — but by then he already saw Apple as a consumer products company).
He never invested in crypto, blockchain, AI.
He only stayed within the few industries he had thoroughly understood — insurance, consumer goods, finance — placing three to five bets per year.
Three to five decisions, supporting a $1 trillion market cap.
What about you? How many decisions did you make last year? How many of them fell outside your circle of competence?

Buffett is 95 now. He's still in that old office in Omaha, still reading 500 pages a day.
He used 60 years to prove the most counterintuitive truth in the world of investing:
Doing nothing is harder than doing everything — and worth far more.

← Previous Contents Next →
Chapter 10 Walmart

He Opened 1,000 Stores, Every One in a Small Town — and All the City Folks Said: "Nobody Shops in Places Like That"

Not going to big cities. Not picking prime commercial districts. Not building a premium brand. Not waging price wars.
It was a veteran from an Arkansas farm, flying his plane in circles over every unheard-of American small town, finding the most remote spot, and then opening the biggest general store there.
He opened over 1,000 of them.
Fifty years later, Walmart is the world's largest retailer, with annual revenue exceeding $600 billion and a market cap over $1 trillion.
Let me tell his story first. After that, we'll talk about why "going where no one else goes" is the smartest business strategy of all.

July 2, 1962. Rogers, Arkansas.
Sam Walton was 44 that year.
He had already been in retail for 17 years. He'd run a Ben Franklin franchise store, done reasonably well.
But there was an idea in his head that every supplier thought was flat-out stupid.
At the time, every retailer — Sears, Kmart, JCPenney — was cramming into cities. Dense populations, heavy foot traffic, high sales per square foot. The logic was flawless.
Sam did the reverse.
He wanted small towns.
Towns of 5,000 people, 10,000 people — places you couldn't even find on a map.
Suppliers told him: "Nobody buys anything in towns like that."
Sam said: "I know. That's why there's no competition."
July 1962. The first Walmart opened in Rogers, Arkansas.
On the sign, a single line: We Sell For Less.
He meant it.
He priced toothpaste 30% lower than the neighboring town.
He later wrote in his own memoir:
"That wasn't a genius strategy. That was my only option. Because I couldn't compete with Kmart on price in the same city — they had the scale advantage of 500 stores."
"But in a small town — you open the first store, and you own 100% of the market."
For eight full years, he opened only 18 stores.
In 1969, Walmart finally incorporated. Eighteen stores. Total revenue: less than $2 million.
Kmart already had over 200 stores by then.
Everyone laughed at Sam — a small-town retailer, destined to stay small forever.
Sam didn't have time to respond.
He bought a used twin-engine plane and flew it himself from town to town.
He'd spot an empty field, land, and walk to the town government office to ask — what's the population here? Any grocery stores nearby? How many hours does it take a truck to drive in supplies?
His criteria were simple: a town under 10,000 people, no grocery store within a 15-mile radius. Good enough.
That's how it was built.

But what turned Walmart into a giant wasn't picking the right towns.
It was something he did that no other retailer was willing to do.
He saved every cent he could, then cut his profit margin to half of what his competitors ran.
Walmart's profit margin was only 3% — the retail industry average was 5 to 6%.
An executive asked: why don't we raise prices just a little? Half a percent. Consumers won't even notice.
Sam said: no. Because the moment we raise prices by half a percent, we start competing with Kmart on the same price level. And our core competitive advantage is — being cheaper than them.
"We're not cutting prices. We're choosing our battlefield."
He poured every penny saved into logistics.
He built his own warehouse network — every warehouse placed within a day's drive, so trucks could complete restocking within a single day.
His drivers knew where every store's back-door key was hidden — unload at 4 a.m., shelves full by 8 a.m., doors open at 9 a.m.
This system started construction in 1980. Today, Walmart operates over 150 mega-distribution centers.
It's not a retail company. It's a logistics company that happens to collect money.
Then in 1983, Sam did something Kmart thought was utterly insane.

He launched a private satellite.
A real satellite — sent into orbital space.
Not because Sam loved space.
Because he needed to monitor every transaction in every store in real time.
From his Bentonville headquarters, he could know in real time — how many bottles of dish soap were sold today in some small town in Ohio. If it fell below the safety stock level, a truck would automatically dispatch for restocking the next morning.
This was 1983 — more than a decade before the internet went mainstream — and no other retailer was doing anything like it.
When asked about Walmart's satellite, Kmart's president said: "I don't understand why that Arkansas hick needs a satellite."
Ten years later, Kmart filed for bankruptcy protection.
Walmart became the number one retailer in the world.
That "hick's" satellite — Kmart never did figure it out.

Sam died in 1992. That year, Walmart's revenue hit $43.8 billion.
Today, Walmart's revenue exceeds $600 billion. Over 10,000 stores worldwide, serving more than 230 million customers every week.
A few months before he died, he received the Presidential Medal of Freedom — accepted it from a wheelchair. After the ceremony, he went back to Bentonville and kept walking stores.

Now stop for a moment.
Have you noticed — every choice Sam made, from start to finish, looked "wrong" to everyone else?
Go to small towns — wrong.
Only do low prices — wrong.
Cut margins to 3% — wrong.
Build your own logistics — wrong.
Launch a satellite — even more wrong.
His standard for every choice wasn't "correct" — it was "advantageous to me."
Others thought it was wrong because they measured Sam by their own standards.
In your current business choices, is there one that everyone calls "wrong," but you yourself know — for you, it's the only right one?
If you can't answer that — then every decision you've made might just be someone else's decision.

Three things.
First: he refused to fight the giants on their home turf.
Small towns didn't matter to Kmart. To Walmart, they were everything.
He used the giants' indifference as his moat — saturating every ignored small market until it was completely absorbed, then moving on to the next.
In your market, are there "small towns Kmart wasn't willing to go to"?
Second: he wasn't cutting prices. He was building a cost structure no competitor could replicate.
Low prices aren't the strategy. A system that allows you to sustain low prices forever — that's the strategy.
Walmart's distribution centers, satellite network, truck fleet — all are parts of that system.
You can cut prices today — but if you don't have 150 warehouses supporting it, cutting prices is just losing money.
After you drop your prices, can your cost structure make it impossible for competitors to follow?
Third: every penny was spent on the business itself — never on anything that just looked good.
Sam drove a 1979 Ford F-150 pickup truck. He got his hair cut at the cheapest barber in town, $5 a cut.
To this day, the headquarters still has cheap tile floors. The CEO of the world's largest retailer has no private office.
Every penny he saved went into the supply chain. The supply chain then lowered prices. Lower prices brought more customers. More customers brought larger purchasing volumes. Larger volumes drove wholesale prices down — the cost flywheel turned itself for 50 years.
Do you have a "flywheel" right now that keeps turning even when you're not pushing it every day?

In the autobiography Sam Walton wrote during the final year of his life, there's a passage:
"If you ask me how to succeed in retail, my answer is — go where others won't go, open stores they don't care about, sell things at costs they can't imagine. By the time they finally realize what you're doing, you've already eaten the entire town."
That twin-propeller plane now hangs in the atrium of the Walmart Museum in Bentonville.
Beneath it, a single line:
"It all started with a small town."

← Previous Contents Next →
Chapter 11 Eli-Lilly

At 38, He Was a Veteran Army Medic. He Put a Label on a Medicine Bottle and Rewrote an Entire Industry

Not inventing a new drug. Not breaking through in chemistry. Not an MD. Not a massive investment.
It was this: after watching too many men die of infection on Civil War battlefields, a retired colonel and pharmacist opened a small pharmacy in Indianapolis. He did one thing no drug maker was doing at the time — he guaranteed that every bottle of medicine was equally effective and equally safe.
One hundred and forty-seven years later, his name is still on the Nasdaq: LLY. Market cap: $887 billion.
Let me tell his story first. After that, we'll talk about why "quality consistency" was once a disruptive innovation.

  1. The American Civil War broke out.
    Eli Lilly was 23. He enlisted. Rose from private to colonel.
    On the battlefield, he saw a killer more terrifying than bullets — scurvy, malaria, infection, dysentery.
    He watched army doctors dispense medicine bought from rural traveling quacks. Some of it worked. Some of it was fake. Most of it — you had no idea what would happen when you opened the bottle and swallowed.
    From that day forward, one conviction was seared into his mind: medicine must be standardized.
    After the war, he briefly tried cotton farming on a plantation — failed. In the Mississippi heat, the laborers ran off. Cotton rotted in the fields. He went bankrupt.
  2. He returned to Indianapolis with almost nothing in his pockets. He made a decision nobody understood — open a pharmacy.
    Not selling medicine made by others. Making it himself.
    He rented a room on the ground floor of a brick building. Hired three assistants.
    The way he worked was completely different from every pharmacy at the time — he didn't mix medicine by intuition. He ran experiments. Batch after batch. Titrated concentrations. Measured active ingredients.
    If a batch didn't meet the standard — destroy the entire batch.
    The standard in the pharmaceutical trade back then was: close enough.
    Lilly's standard was: not good enough means not good enough.
    On every bottle, he printed his own name — Eli Lilly and Company.
    The meaning was: if this medicine doesn't work, come find me.

In the 1880s, he made the boldest decision yet: invest a third of profits into a research laboratory. Every competitor on the street thought he had lost his mind.
Lilly's lab wasn't working on "what sells well right now." They were searching for "what might change medicine five years from now."
He always told his staff one thing: "What we are producing is medicine — not placebos."
1920. Lilly partnered with the University of Toronto and mass-produced the world's first batches of insulin.
Before him, diabetes was a slow-motion death sentence.
After him, diabetes became a metabolic condition you could manage from the pharmacy.
That year, Dr. Banting and Dr. Best won the Nobel Prize for discovering insulin. Lilly didn't care about the prize — he only cared about one thing: "Can the patient buy it?"
He set the price of insulin extremely low — "low enough that no one dies because they can't afford the medicine."
Over the decades that followed, Lilly produced the polio vaccine, erythromycin, and cephalosporins.
Then came 1986 and Prozac — the first generation of antidepressant medications.
You may have heard the name. It became one of the most prescribed psychiatric drugs in human history.
While developing Prozac, Lilly ran into a problem — the molecule's patent protection was only 12 years, and FDA approval typically took 7 to 8. That meant by the time the patent expired, it would have been sold for only four years.
Lilly's CEO made a decision: voluntarily give up some data exclusivity rights and request accelerated FDA review — on the condition that upon approval, they'd have an unchanged five-year exclusive sales window.
The FDA approved. Within five years of launch, Prozac sold over $10 billion.
They didn't use patents to protect profits. They used time — get to market faster, seize the market earlier, and then bake that first-mover advantage into doctors' prescribing habits.

Now stop for a moment.
Have you ever had a moment when you realized your industry had a "universal but never-solved problem"?
Every drug maker before Lilly tolerated "batch inconsistency" — but no one thought of it as a problem. Because that's how everyone did it.
Until a man came back from the battlefield, having seen fake medicine kill people, and decided to spend his life changing it.
That "industry norm" you've grown used to every day — the thing everyone knows about but no one fixes — how much is it worth?

Three things.
First: he didn't make money first and do R&D later. He invested in R&D first, and let profits come find him.
From the day of its founding, Lilly established its own laboratory. In 1880, that lab was the first internal R&D department in the global pharmaceutical industry.
The concept that "a pharmaceutical company should have its own scientists" — Lilly invented it.
Does your company have a department that "no one else in the industry has" — not because you have excess cash, but because you believe it's your moat?
Second: he turned threats into opportunities. But what truly drove Lilly's sustained growth wasn't competitors' mistakes — it was an obsession with new molecules.
Every time Lilly's most profitable drug patents expired — Prozac, Cymbalta, Zyprexa — analysts predicted a profit collapse. And every time, Lilly's next-generation drug was already queued up at the FDA.
If your core product expires next year — do you have the next one already in the queue?
Third: a name is a contract.
"Eli Lilly" printed on the bottle wasn't just a trademark. It was his own name.
It meant — if something went wrong, he couldn't blame "the company." His name was his personal signature.
Would you put your name on your product — not the company name, your own name?
If you wouldn't dare — how much confidence do you really have in what you're making?

Lilly's insulin pricing was so low it bordered on charity. Not because he was a saint. Because he believed one thing: the patient has to stay alive to keep buying your medicine. If you drive a patient into poverty — you've lost a lifetime customer.
This is capitalism at its most intelligent — trading low prices for a lifetime of trust.
That brick pharmacy building in Indianapolis is now the starting point of Lilly's global headquarters.
On the ground floor hangs that label: "If it bears a red Lilly, it is the right quality."
He was selling medicine. But even more, he was signing his own name.

← Previous Contents Next →
Chapter 12 Micron

Three Engineers Built Memory Chips in a Basement. Japanese Giants Drove Them to the Brink of Bankruptcy — and Then They Did One Thing: No Bailout. Keep Building.

Not government assistance. Not patent lawsuits. Not merging to survive.
It was this: in the 1980s, when Japanese DRAM chips swept across the American market like a tsunami, a dozen U.S. chip companies folded one by one. Micron didn't.
It survived. And then it outlasted every Japanese competitor.
Today, its market cap is $829 billion. It is America's last memory chip giant.
Let me tell its story first. After that, we'll talk about why "survival" is a profoundly underrated capability.

  1. Boise, Idaho.
    Three engineers who had quit Mostek — Ward Parkinson, Joe Parkinson, and Dennis Wilson — pooled $300,000. In the basement of a dentist's office, they founded Micron.
    Idaho was famous for potatoes. Not chips.
    When they went to a local bank to apply for a loan, the bank manager asked: what is it you do?
    They said: semiconductor memory chips.
    The manager said: can you grow that on potatoes?
    This wasn't funny. This was Boise in 1978 — 800 miles from Silicon Valley, the nearest chip fab three states away.
    Why did they choose Boise?
    Because land was cheap. Industrial electricity was cheap. And people here wouldn't hop jobs to the competitor next door — because it was potato fields, there was no competitor next door.
    The first three years, they teetered on the edge of bankruptcy every single day. Chip manufacturing isn't software. A single lithography machine cost millions. A single trial fabrication run burned through hundreds of thousands.
    Their first DRAM chip — 64K — finally entered mass production in 1981.
    Then Japan arrived.

Mid-1980s. Fujitsu, Hitachi, Toshiba, NEC, Mitsubishi — the Japanese government orchestrated a national-level counteroffensive in the chip industry. Built factories. Handed out subsidies. Crashed prices.
DRAM prices fell 80% in 18 months.
Not 8%. Eighty percent.
Intel exited the DRAM market. AMD exited. Texas Instruments exited. Motorola exited too.
America's DRAM companies lined up on a death list.
Micron wasn't dead yet — but it was down to its last breath. Its manufacturing costs were far higher than the Japanese producers'.
The Parkinson brothers made a decision — don't exit.
Not because they had better technology. Because they had noticed something no one else was paying attention to:
Japan's DRAM companies were selling below cost. Every chip was a loss.
This was unsustainable. The Japanese government couldn't subsidize forever.
So Micron's strategy wasn't to beat the Japanese manufacturers — it was to outlast them.
They cut every cost that could be cut. Abandoned every R&D fantasy — only DRAM. Only one kind of chip. Make it the cheapest in the world.
And then wait.
How long did they wait? Nearly ten years.
1993. Japan's bubble economy burst. Government subsidies were forced to shrink. Japanese DRAM companies toppled in waves.
That year, Micron posted its financial report — profitable. Over the next five years, it devoured every piece of market share the Japanese left behind.
This story repeated itself in the 2000s. This time, the opponents were South Korea's Samsung and SK Hynix.
2007 to 2008. DRAM prices collapsed again. Every DRAM company was bleeding money. Samsung and Hynix kept expanding production — using losses to starve their competitors to death.
Qimonda went bankrupt. Elpida went bankrupt. Once again, Micron became the only surviving American DRAM company.
Then it surged forward and bought up the factories and equipment of the fallen — at the lowest possible prices. Including, in 2012, the purchase of bankrupt Japanese Elpida for $2.5 billion.
2013. It absorbed all of its competitors' production capacity — and its cost structure was now lower than Samsung's.
Outlast — buy — outlast again — buy again. This is Micron's law of survival.

Now stop for a moment.
Can you, in a price war, refuse to cut prices — first slash your costs to a level no one can undercut — and then wait for your competitors to exit, one by one?
This isn't cowardice. This is stillness defeating motion.
Micron never won a single "technological innovation war." But it won every single "war of endurance."

Three things.
First: not the best technology. The lowest cost.
Micron's core competitive advantage wasn't building better memory. It was driving the cost of the same memory chip to the lowest in the world. They built fabs in Idaho not because they loved potatoes, but because electricity there was a third of Silicon Valley's price. Electricity is one of the largest variable costs in a chip fab.
In your industry, what's the biggest cost? Have you driven it to the lowest in the industry?
Second: buy capacity when others are afraid.
Every time DRAM crashed, Micron bought. Bought equipment. Bought factories. Bought patents. Bought talent — at liquidation prices. The trough of the cycle is when the weak close their doors, and when the strong expand. Micron understood this.
Have you ever bought assets during the most pessimistic cycle your industry has seen?
Third: extreme focus.
From 1978 to today, Micron has only ever done memory. DRAM. NAND. HBM. Not CPUs. Not GPUs. Not mobile phones.
Its competitor Samsung — phones, displays, home appliances, shipbuilding, insurance. Does everything.
And Micron, in one single product category, outworked Samsung.
Right now, has your drive to "do more" diluted the firepower on your most core product?

Micron set out from the potato fields and survived three industry extinction events. It didn't win at the starting line — it won by never collapsing along the way. Surviving matters more than anything else.

← Previous Contents Next →
Chapter 13 JPMorgan-Chase

The Panic of 1907. Every Bank Was Pulling Cash and Running. He Shut the Door of His Study and Did the Opposite: Summoned America's Biggest Bankers and Ordered Them to Pool Their Money and Save the System

Not an order from the Treasury. Not an act of Congress. Not government intervention.
It was a 70-year-old banker, in the study of his Manhattan home, locking dozens of banking titans — who loathed each other — inside a single room, and saying: nobody leaves until we have $25 million.
And then he put up $10 million of it himself.
That day, he stopped a nationwide bank run. The U.S. government watched from the sidelines the entire time.
Let me tell his story first. After that, we'll talk about the highest tier of influence in the business world — making everyone listen to you, because they know you're the only one who can save them all.

John Pierpont Morgan. J.P. Morgan.
Let me introduce him first.
Born in 1837. His father, Junius Morgan, was a top-tier London banker. He wasn't self-made. He had a starting platform.
But he did one thing no trust-fund child could ever pull off.
He leveraged his father's resources to their absolute limit — and then far surpassed his father.
Young Morgan had one very famous habit. He'd walk into a room full of entrepreneurs and, for most of the time, say nothing at all. His eyes would fix on whoever was speaking, like he was appraising an antique.
People said that after being in a room with Morgan, you might not remember a single word he spoke — but you'd remember those eyes, never once leaving you.
He wasn't socializing. He was assessing who was worthy of trust.

Then America entered the railroad age. After 1865, railroads were the nation's largest industry. And its largest chaos. Every stretch of track belonged to a different company. Getting from Chicago to New York might mean switching trains across three different railroad companies, buying three different tickets.
Vicious competition. Every line undercut the next. Two owners built parallel tracks along the same route — just to steal each other's business. Ninety percent of American railroad companies were losing money.
Morgan saw one thing clearly: the railroad industry needed a "consolidator." Not someone who laid track — someone who assembled the pieces according to industrial logic.
He developed a method of consolidation that came to be known as "Morganization."
He'd buy a troubled railroad company, replace the management, restructure the debt — then acquire its competitors, eliminate redundant lines, and unify fares. Return on capital would finally surpass the cost of capital.
Then he pushed for industry consolidation and stability. He decided which route should be operated by whom, and that fares shouldn't rise high enough to crush freight volume — but also shouldn't fall so low that no one could survive. He demanded every company's board give him a management seat — not because he craved power, but because he didn't believe competition could self-regulate without oversight.
He substituted his personal authority for the "price discipline" that the market economy was supposed to provide. At the time, this was more effective than antitrust law.
By 1900, J.P. Morgan controlled one-sixth of all railroad mileage in America. He wasn't stealing the railroad companies' profits — he was eliminating destructive competition so that railroads could earn stable profits for the first time.
And then he said something that railroad history has preserved:
"I don't like competition. Competition destroys value."
That line grates on the ear today. But back then, he was speaking to a fact he could see with his own eyes — standing amid piles of rusted rails.

But what truly carved Morgan's name into American financial history was 1907.
October 1907. The Wall Street Panic began. A trust company — the Knickerbocker Trust — suffered a run. Then panic spread like an electric shock to other trusts, to banks, to the stock market. No one knew which banks were safe — so everyone withdrew cash. Market liquidity evaporated.
There was no central bank. No Federal Reserve. The U.S. government didn't have a single dollar to rescue the market.
Morgan was 70 that year. His doctors had warned him that his heart couldn't withstand too much stress. His daughter stood guard outside his study those days, refusing to let anyone disturb him.
He walked into his library — the marble building on Madison Avenue in Manhattan — and locked the door. His aides summoned New York's most powerful bankers, one by one.
James Stillman (president of National City Bank). George F. Baker. A long line of trust company heads. They didn't like each other. Some had just fought a brutal battle in another market.
Morgan stood beside the fireplace, one hand resting on the mantel. He looked at them and spoke:
"We need $25 million to cover tomorrow morning's clearing house settlement. If we don't produce that money today — 50 brokerage houses will fail tomorrow, then their banks, then yours."
No one responded.
Morgan said: "J.P. Morgan & Company will put up $10 million first. The remaining $15 million — you divide it among yourselves."
Then he locked the door. Put the key in his pocket.
At 2 a.m., the $15 million was assembled. $25 million was injected into the clearing house the next morning. After the market opened, the panic subsided over the next 48 hours. The run stopped.
After the Panic of 1907 ended, the politicians in Washington realized something — this nation's financial order depended entirely on a single man. And that man would not live forever.
Six years later, in 1913, the Federal Reserve was born. Not because economists designed it — but because J.P. Morgan, in 1907, single-handedly performed the function of an entire central bank, and congressmen were terrified by what that meant.

  1. Morgan died in Rome. He was 75. The New York Stock Exchange flew its flag at half-mast. All trading on Wall Street was suspended for two hours.
    His son, J.P. Morgan Jr., inherited Morgan Bank. Twenty years later, it was broken apart — the Glass-Steagall Act of 1933 forcibly separated investment banking from commercial banking. Morgan Bank was split in two: one half — Morgan Guaranty Trust, which took the deposit and lending business, later becoming JPMorgan Chase. The other half — Morgan Stanley, which took the investment banking business.
    J.P. Morgan surely never wanted his company to be dismantled. But the irony is — after the breakup, both Morgans became among the world's most elite financial institutions. JPMorgan Chase's market cap in 2025: $801 billion. Morgan Stanley: $297 billion.
    The man died. But the name "Morgan" is printed on the names of two trillion-dollar-scale banks.

Now stop for a moment.
There's something you might not sense unless I point it out. J.P. Morgan spent his entire life making money without a single business model or technological invention. His core skill was two words: credit judgment. When everyone else was losing their minds, he could judge who was trustworthy and who was lying.
How much of that judgment do you have right now? Are you evaluating people by their resumes — or by their eyes?

Three things.
First: in the greatest panic, be the sole ballast.
When the market panics, no one wants to put up money first. The first to put up money takes the greatest risk — because if you throw your money in and the market still collapses, you lose everything. But Morgan put up $10 million first.
Why? Because he was the only person in that room who understood — if he didn't put up that $10 million, every dollar that followed would turn to worthless paper.
In your market's most panicked moment — have you ever been the first one to jump?
Second: he didn't compete. He consolidated.
Morgan's core business philosophy his entire life was not "beat your competitors." It was "buy all your competitors, and then end the war." Railroads. Steel (he consolidated Carnegie Steel). All the same logic.
You're not competing with rivals. You're paying "the cost of competition." In your industry right now, is there someone who could end everyone's war of attrition — and why isn't it you?
Third: the compound interest of personal credibility.
On that night in 1907, all the money on Wall Street obeyed one man's direction. Not because he was the richest man in the world. Because for the previous 40 years, everything he had done in every crisis could be summed up in one sentence: "Morgan helped, and he didn't use the moment to rob anyone." The compounded interest of the reputation he had accumulated was redeemed in full on a single night in 1907.
The credibility you're accumulating now — can it, at a critical moment, transform into a massive asset?

After J.P. Morgan died, one line of his was preserved: "People ask me how to accumulate wealth. The answer is simple: be fearful when others are greedy, and write checks of credit when others are fearful."

← Previous Contents Next →
Chapter 14 AMD

Intel Beat Him Down for 50 Years. Every Time He Was Nearly Dead — He Got Back Up and Kept Fighting

Not better technology. Not more money. Not a more elite team. Not more patents.
It was this: Intel forever sat on AMD's head. AMD's market share almost never exceeded 25%. Intel spent more on R&D every single year than AMD's entire annual revenue.
But this company never died.
Fifty years. Every time Intel thought it had taken its last breath — it gasped back to life. And by 2025, its market cap reached $712 billion.
Let me tell its story first. After that, we'll talk about why an "unkillable opponent" is sometimes worth more than a "vanquished enemy."

  1. Silicon Valley.
    Jerry Sanders was the director of sales at Fairchild Semiconductor.
    He wore custom-tailored suits — hot pink ones. Drove a Ferrari. Talked fast, could go ten minutes without a breath. Standing among Silicon Valley's plaid-shirted engineers, he looked like a flamingo in a flock of pigeons.
  2. Eight of Fairchild's core engineers resigned en masse and founded Intel.
    Sanders was not among those eight. He was in sales, not technology.
    The night he found out, he sat in his car for a long time. He later said the only feeling he could remember was — being left behind. He called one investor after another. Everyone asked the same thing: where are your engineers? Where is your core technology? Sanders had no answer. He sent over a hundred letters. No response.
    Eventually, he persuaded Arthur Rock, Intel's chairman at the time, and a few other investors to put up several million dollars.
    Then he went out and recruited other Fairchild engineers — the "second-string" ones.
    He said something that was later written into AMD's history:
    "Intel has the best engineers. And us? We have — everyone who was left."
    "But the ones who were left have something Intel doesn't."
    "We have to win. They're unsackable scientists. We're fighters who could be out of work tomorrow. Either fight, or starve."
    This was AMD's DNA. If an Intel engineer produced nothing this year, the salary kept coming next year. If an AMD engineer had nothing to show in six months — the entire group shut down.
    This defined the next 50 years of offensive and defensive warfare.

1982 — the turning point.
IBM wanted to build a PC. They needed a CPU. They went to Intel. Intel gave them the 8088 — but IBM said: we need a second source. A PC can't depend on a single supplier. What if one day Intel can't deliver?
Intel was forced to sign a second-source agreement — licensing the 8086/8088 to AMD.
That license was AMD's lifeline. Sanders cradled it like a newborn baby.
Then came 1986 — Intel reneged.
New CEO Andy Grove's decision: no renewal. The 386 chip license would not go to AMD.
Sanders's first reaction wasn't anger. It was panic.
AMD had invested hundreds of millions in 286/386-compatible chips — if the license was revoked, all of that investment was wiped out.
But Grove didn't budge an inch. Intel was going from "supplier" to "monopolist."
Sanders launched a suicidal counterattack. He took Intel to court. An eight-year legal battle that burned through hundreds of millions in legal fees. In 1994, the California Supreme Court ruled in AMD's favor — the 386 microcode did fall within AMD's licensed scope.
But by then, the 386 was obsolete. The lawsuit was won. The battlefield was gone. Intel had already released the 486, the Pentium, MMX — every step left AMD in the dust.
This victory cost them an entire era.
But once again, Sanders didn't let AMD die.
He took his team and did one thing: reverse-engineer Intel's architecture. 1996: released the K5 — a total failure. Performance lagged badly behind Pentium. 1997: K6 — signs of life. 1999: Athlon.
Athlon was the first CPU in AMD's history to truly outperform Intel's equivalent generation. Not a cheap substitute. Genuinely faster.
At Comdex that year, Sanders held up the Athlon chip on stage, and every engineer in the exhibition hall rose to their feet.
AMD's stock quadrupled within 12 months. It was the brightest moment in AMD's history — it wasn't just fighting Intel head-on, it had won a round.

Then Intel came roaring back.
After the Pentium 4 debacle, Intel developed the Core architecture — and drove AMD into the abyss once again. From 2010 to 2015, AMD's CPUs barely existed. Server market share fell to zero. Overall market share dropped to single digits.
The stock fell below $2 in 2015. Analysts wrote report after report: AMD's debt ratio is too high, its core design talent has walked. This time, it really is going to die.
Lisa Su took over as AMD's CEO in 2014. Her first major decision was to cut all strategic resources for peripheral products and pour every dollar and every engineer into a next-generation architecture codenamed "Zen." The bet: if Zen failed, AMD would exit the CPU business forever.
February 2017. Ryzen launched. This time, it truly stood on equal footing with Intel. 2018: EPYC server chips tore a hole through Intel's wall in the data center space. 2019: TSMC's 7nm process took the lead — for the first time, AMD pulled ahead of Intel in manufacturing process technology.
2022. AMD's market cap — at one moment — surpassed Intel's. A company that had been beaten down for 50 years had, for the first time, overtaken its rival in market valuation.
Lisa Su wrote four words in an all-staff email: "We are back."

Now stop for a moment.
In 50 years, AMD was never the industry number one. Not once. But it never exited the arena either. Could you survive in the shadow of a giant for 10 years, 20 years, 50 years — and then wait for that one, single opening to appear?
Most founders can't. Because most, somewhere along the way, think: "Forget it. Let's do something easier." But neither Sanders nor Lisa Su ever did.

Three things.
First: the survival strategy for second place isn't catching up. It's out-breathing the leader at a single point.
AMD never survived by "being smarter than Intel." It survived by "still running when Intel slowed down to catch its breath." After the 1982 license revocation — it had no technology, no money, no patents — only one option: reverse engineering. It put in five more years of work, and when everyone thought there was no hope left, it delivered the K6.
What place are you in right now? When your opponent pauses to breathe — do you accelerate?
Second: always be the second option beyond the ARM architecture — that's AMD's positioning value.
Google, Amazon, Microsoft — not a single cloud giant wants to be fully locked into Intel's x86. So they "keep" AMD — not because they love AMD, but because AMD must stay alive to counterbalance Intel.
Does your very "existence" hold irreplaceable value for the industry supply chain?
Third: the CEO can change, but the "underdog gene" cannot be lost.
Jerry Sanders was sales. Lisa Su is engineering. Completely different people. But AMD's core has never changed: we are poorer than our opponent — so we must be more ferocious than our opponent.
Have you given your team a culture that says "we win because we have no way back"?

AMD never won the war against Intel. But it has lived with more resilience than Intel. In the business world, sometimes "not losing" is already the strongest way to win.

← Previous Contents Next →
Chapter 15 Exxon-Mobil

1870, Cleveland, Ohio. A 31-Year-Old Bookkeeper Called Everyone Into His Office and Said: From Today On, We Do Only One Thing — Drive Down the Cost of Shipping Every Barrel of Oil

Not finding oil fields. Not inventing new products. Not developing new technology. Not raising capital for expansion.
It was John D. Rockefeller. A stern, silent man who went to the Baptist church every Sunday. He never smoked, never drank, never played cards in his life. There was only one arena where he was "immoral" — he could find profit hidden in numbers better than anyone alive.
He drove the cost of shipping a barrel of oil from $2 down to 50 cents.
Then he controlled 90% of the world's oil refining.
Today, the descendant of Standard Oil — ExxonMobil — has a market cap of $624 billion.
Let me tell his story first. After that, we'll talk about why lowering costs is worth more than raising revenue.

  1. Cleveland.
    John D. Rockefeller was 16. An assistant bookkeeper at a produce trading firm. Annual salary: $300. His job was to stare at every line in the ledger.
    But he had a habit: every day, at the very bottom of the ledger, he'd write a summary line — calculating the day's "profit margin." Nobody had asked him to do this. The boss noticed and stopped him — "Why are you spending time calculating that?"
    John didn't explain. He was simply doing the one thing he was uniquely good at — finding the profit hidden inside the numbers.
    Then 1859. The first oil well in America was drilled in Pennsylvania. The oil rush began. Every day, men burrowed into the black sludge of Pennsylvania's hills — a gusher could make you rich overnight, but the next day the whole thing could explode and burn to ash. It was the wildest speculation of the 19th century.
    Rockefeller watched a lot of men go mad in the oil fields and lose everything. He made a decision that surprised every "smart" person — he wouldn't touch drilling.
    His logic was simple: drilling was gambling. Transportation and refining were fixed-rate businesses — the lower the cost, the steadier the profit.
  2. Rockefeller, now 26, poured all his money into — refining and transportation. He founded the Standard Oil Company.

The first thing he did was negotiate with the railroads. At the time, three railroads in Cleveland were competing to haul oil — all hungry for volume. Rockefeller said to all three: "Consolidate every refiner's volume under me alone, and I'll fill an entire dedicated train for you every single day — but I want a discount."
This practice later acquired a name — rebates. He inserted a secret clause into the contract: not only would his own shipments get a discount — for every barrel his competitors shipped, the railroad would also pay him a "compensation fee."
This meant: every time Rockefeller's competitors shipped a barrel of oil, Rockefeller made extra money from the railroad. His competitors weren't competing against him. They were working for him.
This quickly made it impossible for any competing refinery to turn a profit. Then, one by one, he bought them — at very low prices — and integrated them into Standard Oil's colossal system.
Not because he had better technology. Because he had already seized control of the narrowest choke point — the cost of transportation.

Next, he did something else nobody was doing — he zeroed in on "waste."
19th-century oil refining: gasoline was a waste product, dumped. The only refined product that mattered then was kerosene — used for lamps. Gasoline was too volatile, nobody wanted it. Out of every barrel refined, only 60% became sellable kerosene. Everything else was poured into rivers or burned off into the sky.
Rockefeller found chemists — he built a laboratory in every plant — and asked: can we turn this waste into money?
His chemists extracted paraffin wax from the waste (later used for candles), petroleum jelly, asphalt, and lubricating oils — and then, using the chemical processes they discovered, transformed gasoline into industrial fuel, arriving just in time for the internal combustion engine age.
He wasn't drilling for oil. He was squeezing more products out of every single barrel. What other people dumped as trash became his most profitable byproducts. He fundamentally raised the resource utilization rate of the entire refining industry — and, of course, in the process, annihilated every competitor who couldn't do the same.
By the 1890s, Standard Oil controlled 90% of America's refining capacity and a vast portion of its transportation pipelines. Rockefeller became America's first billionaire. Then 1911 — the U.S. Supreme Court ruled Standard Oil in violation of antitrust law and ordered its breakup.
That breakup of the century created a string of companies — Standard Oil of New York (later Mobil), Standard Oil of New Jersey (later Exxon), Standard Oil of California (later Chevron). And a whole list of companies you've heard of but didn't know were Standard Oil's offspring — today's Marathon Oil, Amoco (absorbed by BP), Conoco, Sohio.
After Rockefeller was "dismembered" from his own company, he didn't grumble in public. He woke every morning at 7 a.m., put on the same old black coat, and walked into church. Then he sat in his garden and watched the birds. Over the rest of his life, he gave away $500 million — built the University of Chicago, Rockefeller University, the Rockefeller Foundation, and funded the team that discovered the yellow fever vaccine.
He applied the most calculating mind of a lifetime to making money. And then he used that money to become one of the greatest philanthropists in history.

Now stop for a moment.
Have you ever calculated — what is the single biggest source of waste in your industry? A customer who doesn't renew — you treat it as "normal churn," but what if you turned it into a new product? That 15% loss in a business process that everyone's been ignoring for years — if you cut it, how much would it lower your costs?
Rockefeller wasn't selling oil. He was taking the most chaotic industry in the world and quantifying every piece of chaos with a slide rule and a ledger. Your people can use a slide rule too.

Three things.
First: control the choke point.
He didn't control the oil wells. He controlled the transportation pipelines and refining capacity — the two things every oil driller had no choice but to pass through. Whoever controls the intersection collects the toll from the entire industry.
In your industry, which link in the chain is unavoidable for every competitor — and have you built a toll booth next to it?
Second: find money in waste.
When others were dumping gasoline as garbage, Rockefeller was already making three new products out of waste — and every one of them was pure profit. He didn't make money by raising prices — he made money from "things nobody else saw as valuable."
In your business, is there a "byproduct" being disposed of as trash — that could be your next profit pillar?
Third: the ultimate dignity of the monopolist — he cloaked every unfair competitive tactic in the mantle of efficiency.
Rockefeller's rebates and his methods of encircling and crushing rivals would be deeply problematic by today's standards. But the other side of the coin is equally true — he really did drive the price of kerosene lamp oil from 58 cents a gallon down to 8 cents, giving millions of American families access to affordable lighting for the first time.
He didn't rely on tactics alone — he relied on tactics plus efficiency. A company with tactics but no efficiency, once broken apart, dies. Rockefeller was broken into 34 pieces — and every piece became a giant.
If your company were forcibly broken up tomorrow — could every piece survive independently, and thrive?

Rockefeller died at 98. He refused every interview request his entire life. The only time he wrote about himself was in his diary: "I'm not betting on luck. I'm doing the math."

← Previous Contents Next →
Chapter 16 Visa

1958. A Bank Mailed 60,000 Credit Cards to Every Resident of a California Town. No One Applied. No One Activated. Every Cardholder Instantly Became Something Else — a Victim of Fraud

This is how the credit card was born. A violent, chaotic, near-miscarriage. Then a man named Dee Hock arrived. The architect of Visa — a man whose thinking ran opposite to everyone else on earth.
He took a piece of plastic that couldn't turn a profit and transformed it into the largest payment network in human history. Today, Visa's market cap is $620 billion. The transactions it processes every day equal the GDP of a mid-sized country.
Let me tell its story first. After that, we'll talk about why Dee Hock believed the two stupidest words in leadership are "take control."

  1. Bank of America ran a marketing experiment.
    They mailed 60,000 credit cards — called BankAmericards — directly into every household mailbox in Fresno, California. No application required. No credit check. The cards were already activated. You could take one and swipe.
    The result? Within two months, the entire city of Fresno had turned into a crime paradise — 20% fraud rate, 20% delinquency rate. The bank lost roughly $20 million on the experiment (equivalent to over $200 million today).
    The Bank of America CEO summoned every senior executive: "Deal with this. You can shut it down too."
    That's when Dee Hock walked in.

Dee Hock's resume was deeply atypical. He had no college degree. He had worked as a wool sorter, a dry-cleaning manager, a bucket hauler. He didn't join a small bank in Washington State until his early thirties. He was later drawn to the chaos inside Bank of America's credit card division — because "chaos means opportunity."
When Bank of America was ready to abandon BankAmericard, Dee Hock made a suggestion that drew roaring laughter from the room:
"Don't shut it down. Give it away."
"Spin it out of Bank of America — turn it into an independent, nonprofit alliance collectively owned by all the issuing banks. Our job isn't to own it — it's to make it capable of moving between every bank."
In other words — Visa should not be owned by anyone. Visa should be a "set of rules" collectively owned by all banks.
Five seconds of silence in the conference room. Then someone laughed out loud: "You want us to give up control of this business?"
Dee Hock nodded.

He turned Visa into the most unique thing in business history — a "chaordic organization."
It wasn't a company. It wasn't a government agency. It wasn't a cooperative. It was a "membership alliance." Every issuing bank — large or small — was a member and shareholder of the Visa network. Visa itself issued no credit cards. Loaned no money. Had no direct contact with customers. It did exactly one thing: maintain a set of rules. When someone in Tokyo swiped a Visa card issued by a New York bank, Visa's system completed, within a single second: authentication, currency conversion, clearing, and settlement.
When you stop and think about it — this is effectively the most complex cross-currency, cross-border, real-time risk management network on the planet. And for over 30 years, it was run by a nonprofit membership alliance.
But this model also contained its own contradiction. In 2007, because the network had become so enormously valuable in the public markets, competition among member banks began to tear apart its internal governance — some banks believed Visa should remain a neutral, nonprofit pipeline, while others wanted to turn it into a for-profit company to unlock shareholder value. Ultimately, in 2008, Visa restructured and went public — becoming the for-profit company we know today (Mastercard went through the same restructuring and IPO the same year).
Though the form changed, Dee Hock's core philosophy of decentralization still permeates Visa's DNA — it is still nothing more than an "end-to-end but frictionless in the middle" protocol layer. It doesn't lend you money. It simply ensures that at the exact moment you borrow, every bank on the entire planet has confirmed your identity within the same single second.

Now stop for a moment.
Have you ever had a moment — where you "gave away" your business? Not sold it. Let all stakeholders co-own it — let them all become your "nodes" — and then you become irreplaceable infrastructure?
The moment Dee Hock surrendered control — that's when Visa truly became Visa.

Three things.
First: don't sell a product. Establish a protocol.
Visa's core is a protocol: one end swipes, the other end confirms, clearing happens in the middle. Any bank capable of abiding by this set of rules — can join the network. Then the network effect amplifies itself — the more people who join, the greater the value every participant receives. Because the more readily merchants accept this card, the more consumers want to hold it; and conversely, the more consumers hold it, the more merchants are compelled to accept it.
Can your business become — not a product, but a set of protocols that every peer in the industry is willing to follow?
Second: hide the profit — let every link in the value chain make money.
Visa doesn't do lending — so the issuing banks make money. Visa doesn't sell anything — so the merchants make money. On its own core clearing and settlement layer, Visa barely breaks even, collecting only a tiny interchange fee. Then, when the network grows to billions of cards, that tiny "skim" becomes tens of billions of dollars in annual revenue.
Can you give the profit to others — and take a tiny percentage of the total volume for yourself?
Third: chaos is a gift.
If that 1958 Fresno experiment hadn't lost $20 million — Bank of America would never have wanted to abandon the business. Dee Hock would never have gotten the chance to pick up that piece of garbage.
The biggest mess on your hands right now — is there a Visa buried inside it?

The year Dee Hock retired — 1991 — Visa processed over $300 billion in global transactions. No one remembers the name of the marketing VP who mailed out those 60,000 cards. But everyone remembers Dee Hock — the man who said "give up control."
He wrote a book called Birth of the Chaordic Age. In it, he wrote a line that Silicon Valley quotes to this day: "The more complex a system becomes, the only way to manage it is — to let it manage itself."

← Previous Contents Next →
Chapter 17 Intel

Eight Engineers Quit En Masse and Founded Intel. The Toughest Among Them Later Wrote a Book Called Only the Paranoid Survive

Not technical genius. Not a patent portfolio. Not capital backing. Not a window of opportunity.
It was Andy Grove — a Hungarian Jewish refugee who, when he arrived in America at age 20, couldn't speak a single word of English. Then, within 30 years, he led a group of the world's most brilliant engineers to do something profoundly counterintuitive: at the very peak, dismantle his own core business with his own hands and start over.
In 1985, he asked Gordon Moore a question: "If we were kicked out by the board and a new CEO came in, what would he do?"
Moore thought for two seconds. "Exit the memory chip market."
Grove said: "Then why don't we walk out that door ourselves, come back in, pretend we're the new CEO, and do exactly that?"
And they actually did it.
Intel exited DRAM. Went all-in on CPUs.
And then humanity got the personal computer age.
Let me tell his story first. After that, we'll talk about why "killing yourself" is the hardest CEO decision of all.

Andy Grove was born in 1936 in Budapest, Hungary. His childhood name was András Gróf.
His father was a Jewish dairy farm owner. In 1941, German forces occupied Hungary. His father was conscripted into a labor camp. In 1944, the Germans packed his mother and a crowd of Jews into a cattle car bound for an extermination camp. She survived — because at a transfer station, she took a different path.
Andy and his mother hid in a basement through the final months of the war. He was 8 years old.
After the war, his father finally returned from the labor camp, reduced to a skeleton.
1956. The Hungarian Revolution erupted. Soviet tanks rolled into Budapest. Twenty-year-old Andy, in the dead of night, with a few friends, crawled through farmland and crossed the border on foot into Austria. He had the clothes on his back and a few pieces of bread.
Less than a year later, he arrived in New York. No English. No family. Less than $50 in his pocket.
He enrolled at City College of New York for his first semester — washing dishes while learning English — and then, in 1960, earned his Ph.D. in chemical engineering from UC Berkeley.
1968. He joined a new company — Intel. He wasn't a founder (the founding trio was Gordon Moore, Robert Noyce, and Andy Grove — technically Grove was employee number one, and he was called the third founder). The company's first products were memory chips — DRAM.
Throughout the 1970s, Intel was the king of memory chips. Without equal.
Grove was the director of engineering operations. His management style — how to put this — was very un-Silicon Valley.
His creed: five minutes late, write a written self-criticism. He would post a clock at the factory gate and lock the doors at 8 a.m. sharp. Everyone who arrived after the doors locked — name written in a notebook. If you didn't work hard, he'd reassign you to a punitive shift called "Scotland" — midnight to 8 a.m., seven days a week. Those who couldn't endure quit on their own.
But it was this very man who, in 1985, saw through to one truth: Intel could no longer make memory chips. The Japanese were destroying the market.

Those months in 1985. Every day when Grove walked into the Intel factory, the air carried a mixed smell of acid etch solution and fear.
Orders were shrinking. Japanese competitors were pricing their chips 30% below Intel's manufacturing cost. The DRAM division was bleeding away a third of Intel's revenue.
Using his "revolving office" approach, he asked Gordon Moore that classic question — if we were kicked out, what would the new CEO do upon walking in.
Moore said: abandon memory. Do microprocessors.
Grove said: then let's walk out this door — pretend to walk out — and then walk back in. And be that new CEO.
This decision had the support of one of Intel's soul figures — Noyce. Noyce was one of the old-school co-inventors of the integrated circuit, and the gentlest and most beloved of the three. He didn't persuade people with logic — he was the one no engineer wanted to disappoint. When Noyce quietly stood up and signaled that "the pivot is right," Grove swept away the last internal resistance.
Then all of Intel's resources — R&D, manufacturing, sales — were redirected entirely. 1985 is remembered as the year Intel abandoned memory chips. The 386 processor launched that year.
What followed unfolded like a row of dominoes —
Compaq used the 386 to build the first IBM-compatible PC that wasn't made by IBM. IBM's licensing wall was breached. PC compatibles flooded the market like a tide. Every compatible machine needed an Intel CPU. 386, 486, Pentium, Pentium Pro, Core — "Intel Inside."
From that act of self-revolution in 1985 onward, Intel went from the king of memory to the heart of the computer. For 30 years, not a single personal computer of significance existed without Intel inside.

But do you know the second half of the story?
2010. The first iPhone chip wasn't Intel's. It was ARM.
Intel had the opportunity to buy the ARM architecture at that time. And decided they didn't need it.
Then — in the very moment when Grove's proudest instinct, the "pivot instinct," failed — the entire mobile internet rolled past Intel. Not a single grain of it landed in their pocket.
After 2010, Intel tried three times to break into the mobile chip market. Three failures. Not because the technology was inferior — because Intel had grown so accustomed to the entire world revolving around the x86 architecture that it refused to abandon its high-margin, high-performance CPU model in order to adapt to the low-power world of mobile chips.
So much so that by 2023 to 2025, people had already started saying: the company that once understood "killing itself" better than anyone is now being killed by others.
This is the cruelest reversal in Intel's story. The enterprise that invented the methodology of eating itself — later became the giant most afraid to take another bite of itself.

Now stop for a moment.
When was the last time you "killed yourself"? Not because someone forced you. You, on your own initiative, while all your KPIs still looked great, voluntarily gave up a profitable old business to bet on a new direction that wasn't yet making money.
If you can't answer — you're just waiting for someone else to come and kill you.

Three things.
First: the key to that self-revolution in 1985 — wasn't that DRAM wasn't making money. It was that Noyce, Moore, and Grove — all three — were simultaneously willing to bet the entire company on a new track.
One CEO can pivot. But can an entire founding team pivot collectively? When the old business is still generating revenue, are your partners willing to burn it down with you?
Second: Grove's theory of the "strategic inflection point."
All of his later methodologies condense into this single concept: every company will encounter, several times in its life, a "strategic inflection point" — a point after which every past success formula becomes wrong. If you can't perceive it before everyone else, your company will become the most expensive fossil in the museum.
In your industry right now — is that "inflection point" already at the door? Do you feel it?
Third: don't let your most profitable product become your future coffin.
Intel's high-margin x86 — made it forever unable to truly embrace the low-power mobile world. High margin = high resistance. Is the most profitable piece of your company blocking you from entering the next track?

Andy Grove died in 2016. His most famous line is — "Only the paranoid survive."
He wasn't crafting a clever phrase. He wrote it with his life.
A man who escaped Nazi concentration camps and Soviet tanks — he knew: if you don't live every day believing someone is trying to kill you — one day, someone really will come and kill you.

← Previous Contents Next →
Chapter 18 Johnson-Johnson

148 Years Ago, He Opened a Pharmacy. His Secret Weapon Wasn't Medicine — It Was One Principle That Never Changed: Patients First

Not a breakthrough in R&D. Not a patent monopoly. Not marketing genius. Not financial engineering.
It was this: 11 years after the American Civil War ended, a veteran army medic opened a pharmacy in Indianapolis and hung a sign at the door — "Serve the patient first. Consider profit second."
He wrote this sentence into the company's constitution. His descendants have executed it for 148 years.
J&J's market cap today exceeds $540 billion.
Let me tell his story first. After that, we'll talk about why a value can become the most valuable business asset of all.

  1. Indianapolis.
    Robert Wood Johnson was 31.
    He had fought in the Civil War and served as a medical assistant on the battlefield. He saw that too many men weren't killed by bullets — they were killed by bacteria.
    From wound infections. Because there were no antiseptics.
    Eleven years after the war ended, he did something every pharmacist of the era thought had no future.
    He embraced Sir Joseph Lister's theory of antiseptic surgery.
    Lister said: before performing surgery, doctors should wash their hands and sterilize their instruments with heat.
    Sounds basic now. But in the 1870s, this was a theory mocked by surgeons everywhere.
    "Germs? How can invisible things kill people?"
    Robert Wood Johnson chose to believe. Together with his two brothers, he founded Johnson & Johnson.
    Their first product — sterile surgical dressings. Sealed in sterile packaging, delivered to every hospital.
    Death rates fell from 50% to 5% in that era. He was one of the driving forces behind that surgical revolution.

But what truly turned J&J into a legend wasn't the innovation of 1876.
It was a single sheet of paper in 1943.
1943. On the eve of the company going public, Robert Wood Johnson II — the founder's son, then chairman — wrote a single page by hand. Not long. Four paragraphs.
He called it "Our Credo."
First paragraph: "We believe our first responsibility is to the doctors, nurses, and patients, to mothers and fathers and all others who use our products."
Second paragraph: to employees.
Third paragraph: to communities and the environment.
Fourth paragraph: to shareholders.
"When we operate according to these principles, the stockholders should realize a fair return."
Note the order — shareholders in fourth place. Not first.
And every CEO since has nailed that sheet of paper to the wall of every J&J office around the world.
The moment that truly tested those words erupted 38 years later.

September 1982, Chicago.
Seven people died after taking Tylenol, a J&J brand.
It wasn't the drug itself. Someone had gone into several pharmacies in the Chicago area, secretly pried open Tylenol capsules, stuffed cyanide inside, resealed them, and placed them back on the shelves.
It was a random, sociopathic act of poisoning. Completely unrelated to J&J's production line.
But the dead were Tylenol customers.
The last Wednesday of that month, J&J CEO James Burke opened the copy of "Our Credo."
He didn't summon the legal team to discuss liability. He didn't spend one minute analyzing who bore more fault.
He did one thing: recall every Tylenol product from every shelf in America.
Not just the Chicago area. All of America. 31 million bottles. All pulled.
Then he publicized a 1-800 hotline number. Told everyone who had bought Tylenol — bring it back. Full refund.
The cost of the recall at the time: over $100 million.
Wall Street analysts called him insane: "Your legal liability is almost zero. Why are you spending $100 million?"
Burke's answer: "Because of the first line of the Credo."
After that recall, Tylenol was back on shelves within three months. With triple tamper-resistant packaging — this packaging later became the FDA's industry standard.
Within a year, Tylenol's market share had recovered to pre-poisoning levels.
J&J did not die in that crisis. It emerged stronger because it had "done the right thing."

Now stop for a moment.
Here is a question every founder, every CEO should be able to answer instantly:
If tomorrow, something went wrong with your product — not your fault, but a customer died —
Who would your first call be to?
Legal? PR? Your investors?
Or the most direct people — the customer, and their family?
If you can't produce the correct sequence in under two seconds — you need to figure this out today.

I've studied J&J for a long time. Three things.
First: the Credo wasn't written to be looked at. It was written to make decisions in moments of crisis.
Every crisis J&J has faced over the last 40 years — Tylenol, contact lenses, talcum powder — in reality, the path to resolution has always been the same: return to the Credo, and play your cards in order of priority.
When your decision-making framework was written down 80 years ago, your speed in a crisis will be ten times faster than any competitor's, because you don't need to figure out in the moment "what should we do this time?"
Does your company have a written-down priority framework for decisions — not inside the CEO's head, but hanging in the lobby, something every employee can recite from memory?
Second: putting shareholders in fourth place turns out to be the most beneficial thing for shareholders.
J&J is a "counterintuitive case study." When you don't put shareholder interests first, they actually end up making the most money — because prioritizing customers and employees creates enduring brand trust and talent density. Trust becomes pricing power. Talent density becomes innovation velocity.
And those two things, in turn, drive profits higher. After the 1982 Tylenol crisis, total shareholder returns went up, not down.
In your pursuit of "making money," have you been skipping steps in accumulating trust, only to lock a ceiling on your profit potential?
Third: "do no evil" is backed by an entire system of processes, not one person's good heart.
J&J's Credo Challenge — management teams from every global branch are called into a conference room each year, and they go through the four paragraphs of the Credo, line by line, discussing: in this year's decisions, where did we deviate from the Credo?
This is not a philosophy salon. This is a KPI review. Failed it? Deduct it from the year-end bonus.
Have your "values" entered your performance system — or are they still living only in the copy on your website?

Robert Wood Johnson didn't live to see the year of the Tylenol recall. But those four paragraphs he wrote made the most important business decision of that era on his behalf.
The first paragraph he wrote in 1943 saved Tylenol in 1982. And in 2026, it is still protecting this company.
The true legacy a founder leaves behind isn't a product. It's a principle.

← Previous Contents Next →
Chapter 19 Oracle

At Age 35, He Had No Money. At a Company Nobody Had Heard Of, He Used an IBM Paper Tossed by the Roadside to Build a Database — Then He Chased His Customers Down One by One, Calling Them Thieves If They Wouldn't Pay

Not a technology breakthrough. Not a product innovation. Not a market tailwind. Not venture capital blitz-scaling.
It was him. Larry Ellison. A Russian-Jewish adoptee from Chicago. College dropout. Rejected by every respectable company. He spoke like a machine gun and laughed like a pirate. And over 45 years, he built Oracle into a $529 billion empire.
Let me tell his story first. After that, we'll talk about why shameless confidence is sometimes worth ten times more than technical ability.

1944, New York.
A 19-year-old unmarried girl gave her newborn son to her aunt and uncle in Chicago. They named the baby Larry.
His aunt's family wasn't wealthy. But his uncle was a man of vision. He often told Larry one thing: "You have nothing to be afraid of. You are meant for great things."
At South Shore High School in Chicago, Larry was a math prodigy — and the most insufferable troublemaker in the building. The universities he attended — University of Illinois, University of Chicago, Northwestern — none of them ever put that piece of paper called a diploma in an envelope for him. He couldn't be bothered.
He couldn't be bothered with anyone's rules.
At 22, he arrived in California. A few dozen dollars in his pocket. He worked as a programmer, as an insurance company clerk, got fired several times. In one interview, the interviewer asked: "What do you think your weaknesses are?" He answered: "I'm too smart."
He didn't get that job.
In 1977, while working at Ampex, he read a publicly published paper from IBM Research: System R: Relational Database System Design. He read it five times. Then he did something that would change the database industry — he started writing code.
He believed the "relational database" would be the future of business computing. Not cramming data into dead tables, but turning every "relationship" between data points into queryable logic. At the time, this idea was deeply unfashionable.
He took two colleagues — Bob Miner and Ed Oates — left Ampex, and founded "Software Development Laboratories." Startup capital: $2,000.
The company was later renamed Relational Software Inc. — and later still, renamed Oracle.

The first big order — the CIA. They needed a relational database capable of handling vast volumes of intelligence records. Ellison said yes without hesitation, even though Oracle version 1 — as they well knew — was riddled with bugs and plagued by unstable connections.
The CIA's payment kept them alive. Then came version 2. He jumped the version number from 1 to 3 — deliberately skipped 2, because he figured "nobody buys version 2 of anything."
That was Ellison's sales philosophy — not selling a product. Selling confidence. He believed he was the best in the world first, and then the customers believed it too.
In the 1980s, Oracle's competitors — Ingres, Informix, Sybase — were often far more technically mature and reliable than Oracle. But those companies were led by university professors or humble engineers. None of them could match Ellison's ferocious sales intensity.
His sales force was known throughout the industry as the "Oracle Wolfpack." The alpha wolf would lead the team to camp out in a hotel lobby across from the client's office, making dozens of calls a day. His message to the salespeople was unambiguous: close the deal — method doesn't matter.
If a customer reported that their Oracle database had crashed, Ellison would personally call to apologize, then say: "Our new version has fixed everything." Then he'd return to headquarters and roar at the engineers: "Fix the bugs before the customers find us!"
He was doing two things simultaneously: externally — hyping a glass half-full into a vast ocean; internally — frantically filling that ocean to the brim. This dual-engine cycle ran for 45 years.

In the 1990s, Oracle met its first existential crisis. Version 6 was a quality disaster — the entire suite was profoundly unstable, and major clients fled en masse. Informix surpassed Oracle on both performance and SQL standards. In 1990, Oracle's revenue nearly flatlined.
Ellison's response — cut all non-core businesses. Database only. Then he tied every engineer's bonus directly to "the number of SQL standard compliance tests passed." At the all-hands meeting, he said exactly one sentence: "If another customer loses data — I will personally come to your desk."
Then Oracle 7 was released in 1992 — a landmark in database history. Transparent distributed transactions. More reliable than most customers had dared to expect. It lifted the database from single-server machines into the enterprise client-server era. Soon all of Wall Street and every Fortune 500 company was lining up to sign Oracle contracts.
In 2005-2010, he continued to be known for his aggression. His acquisition roadmap: buy the biggest company in every competitive domain — PeopleSoft, Siebel, Hyperion, BEA, and later Sun Microsystems and MySQL. Mostly software-related rather than chip-related. He didn't chase slowly. He bought the entire mountain.
Each of these acquisitions was dismissed by analysts the day it was announced. And every single time — he locked in the acquired product's customer base, surrounded them with the Oracle database, raised prices, cross-sold — and the entire portfolio started printing money for him.
By the 2010s, Oracle had transformed from a database into the full-stack backbone of enterprise IT.
Then the cloud era arrived — and AWS and Salesforce punched him hard. 2015 through 2019 were the years Oracle was questioned most — it was too slow, he was too old.
But Ellison's reaction was identical to the last crisis. He pulled all resources out of the traditional licensing model and forcibly wrenched them into a cloud subscription model. From 2022 to 2025, Oracle's cloud revenue began to accelerate. Ellison looked at the analysts, smiled, and said — "I'm alive again."

Stop here for a moment.
When you were young, if people kept telling you: don't boast, you're not good enough yet — did you believe them? Or were you like Ellison — talking big first, then chasing your own words to force yourself to catch up?
The world will always reward those who "throw their hat over the wall first." Because ordinary people see the wall and turn back. But people like Ellison — after they've run their mouth, the only option left is to climb over it.

Three things.
First: Sales can beat Engineering — as long as Sales forces Engineering to run faster than the competition.
Most engineers believe "the best product will naturally get bought." Wrong. The best product, if unheard, doesn't exist.
Between "selling" and "building," have you placed all your faith in "building" while underestimating "selling"?
Second: Always skip version numbers.
Ellison skipped Oracle version 2, jumping straight from 1 to 3. Not because the technology had leapfrogged — because he knew customers only buy "the latest." For your next product, can you skip those intermediary versions nobody notices and go straight to a major version that everyone has to notice?
Third: Don't wait to be "perfectly ready." Land the customer first, then let the deadline burn your internal team alive.
Oracle's first CIA contract was built on an incomplete product. And that contract became a furnace — forging a working product out of the fire.
Have you given yourself a "non-refundable deadline"? If not — your team will forever be preparing, never delivering.

Larry Ellison is 81 this year. He's still Oracle's CTO. His yacht is longer than the Eiffel Tower. He spent hundreds of millions building a replica of a Japanese imperial palace. He's raced planes and nearly died more than once.
He doesn't need to be liked. He doesn't need to be respected. He only needs — to win.

← Previous Contents Next →
Chapter 20 Costco

He Slashed Retail Gross Margin to 14% — Then Told Every Board Member Who Wanted to Raise It: Whoever Raises It, Leaves

Not launching a premium line. Not adding more SKUs. Not running ads. Not opening more formats.
It was Jim Sinegal. A man who grew up pushing shopping carts in a supermarket. He had one obsession his entire life — cost price plus 14% is the selling price. Not a penny more.
In his 30 years running Costco, the average gross margin in American retail climbed from 30% to 40%. Costco's gross margin stayed exactly where it was — 14%.
And Costco became the third-largest retailer in the world, with a market cap of $453 billion.
Let me tell his story first. After that, we'll talk about why "refusing to earn more" is sometimes the smartest way to earn.

1982, Seattle.
Jeffrey Brotman called Jim Sinegal. Brotman was a lawyer by training, and he'd been searching for a retail veteran to partner with. Sinegal had spent most of his life in retail, working his way up from pushing carts on the supermarket floor to the executive ranks — he had seen every trick in the book for "making customers spend money they shouldn't have to." His whole life, he'd been thinking about the same question: could you build a supermarket that "never overcharges a customer by a single cent"?
The two men clicked instantly.
September 1983, the first Costco opened in Seattle.
Walk inside, and what you saw wasn't decor, wasn't displays, wasn't sales assistants. It was bare concrete floors, industrial shelving, pallets stacked to the ceiling in giant packaging. A few thousand SKUs. A typical supermarket carries around 30,000 to 40,000 SKUs. Costco typically carries only 3,700 to 4,000.
Why?
Because every category gets only one brand — the best one. No "three tiers of toothpaste for you to compare." If you want cheap, you buy Kirkland — the private label. If you want a name brand, you buy Colgate — but it has to be the Costco-sized mega-pack. The more you buy of a single SKU — the cheaper the wholesale price — the lower the price versus everyone else.
This isn't simple retailing. It's a deliberate arms-race logic — you get more for less money, but you have to buy it by the case, by the pallet. You bought 48 rolls of toilet paper — Costco doesn't spend extra labor on warehousing and restocking — because that 48-roll pack moves through the system far faster than selling 6-roll packs eight times over. This isn't punishing the customer — it's passing the compounding interest of operational efficiency back to the customer.

And Costco's deepest secret — is that 14%.
Sinegal wrote a memo to his executives and board. It was short: "We do not cheat our members on price."
Later, someone asked him: if customers are willing to pay 20% more, why not float the margin up to 16%?
Sinegal's answer: "If we raise it 2 points today — two years from now, we'll be like every other retailer, raising it 4 points, then 6 points — and one day you'll realize you're no different from anyone else."
"Raising prices does exactly one thing: it lets you stop needing to get better."
Throughout his entire tenure, the Costco board seriously discussed raising the gross margin at least twice. In those boardroom minutes, Sinegal's answer was always the same three words: "Not over my dead body."

2009, the financial crisis. Every retailer was under crushing pressure. Walmart launched massive promotions. Target slashed prices to clear inventory.
What did Costco do?
It didn't cut prices.
Because the prices were already the lowest possible.
That year, Costco's membership renewal rate was 87%. Not rising. Steady. Steady at 87% while the economy was collapsing. Because in a crisis, everyone is searching for the best way to spend their money — and they discovered that Costco, without running a single promotion, was already that "best way."
A deeper secret: Costco doesn't make its money selling merchandise. It makes its money from membership fees. Costco's annual fee income is roughly equal to its entire annual net profit. Its retail profit barely covers operations — margins so low they border on charity. When a member walks up to the register — they are redeeming the membership fee they already paid.
All of its profit — comes from the people signing up for a card at the front door. Not from a bag of chips on the shelf.
When your business model isn't "sell more stuff to pocket the spread" but instead "customers pay upfront every year," your interests and your customers' interests are no longer adversarial — they are perfectly aligned. You have to save your members money every single day, otherwise they don't renew the following year.
Wall Street analysts never liked this model — gross margins too low. Costco never responded.

Stop here for a moment.
Do you have a pricing strategy — one that can hold gross margin absolutely motionless for 30 years — no matter whether the economy is soaring or sinking? If your industry's average gross margin is 30%, would you dare to take only 14% — then use radically transparent low prices to build a lifetime of trust with your members, making your money somewhere else?
If you wouldn't dare — is your "moat" really just a decorative wall built on empty ground that others simply haven't bothered to occupy yet?

Three things.
First: Don't make money selling products — collect rent from membership trust.
Costco's profit structure is unlike that of any traditional retailer. It's not happier when you spend more — it's happier when you save more. Because only when you feel you've won do you renew your card.
Is your profit built on your customers' satisfaction at "saving more" — rather than on the depletion of "buying more"?
Second: SKUs are the enemy.
Costco averages 3,700 SKUs. Walmart's typical supercenter carries over 100,000. Sinegal discovered a phenomenon: for every additional SKU, the complexity of restocking and the labor cost both grow non-linearly. Eliminating 90% of choices isn't about reducing customer options — it's about making every remaining choice supremely optimal.
Your current product count — does it make it easier for customers to choose, or harder?
Third: Never raising prices is a strategy, not a charity.
The 14% gross margin wasn't Sinegal being kind. It was him seeing that — break the rule once today, and you'll break it again tomorrow. Today's 2-point hike is a "special circumstance," and next year there'll be another "special circumstance." Only by never yielding can you make sure that everyone — buyers, store managers, members — knows forever what a Costco price actually means.
Do you have a "never compromise" principle — one that every person on your team can recite from memory?

Sinegal retired in 2012. Before stepping down, at his final employee town hall, he said: "The smartest business decision I ever made in my life — was not fighting my customers for money. Every retailer wanted higher profits, and I wanted more members. In the end, my members became worth more than all my competitors' profits — combined."

← Previous Contents Next →
Chapter 21 Mastercard

In 1966, a handful of banks sat down together and decided to build an "anti-Visa alliance." Today, it's worth $445 billion.

Not a technological breakthrough. Not a disruptive business model. Not a first-mover advantage. Not brute-force capital.
Here's what happened: after Visa's BankAmericard had already monopolized the national credit card network, a group of excluded banks — from Buffalo, Pittsburgh, Kentucky — gathered in Buffalo, New York, for a meeting. They had one goal: build a second network. Before them, Visa was the only choice. After them — merchants had two choices, consumers had a comparison, and the credit card network became a duopoly.
Mastercard's market cap today: $445.7 billion.
Let me tell this story first. After that, we'll talk about why a "coalition of second place" sometimes outruns a "dictatorship of first place."

1966. Buffalo, New York.
On a frigid winter night, the CEOs of more than a dozen small and mid-sized regional banks flew into this snow-buried old industrial city. They weren't here on vacation. They were here to discuss how to survive.
Bank of America's BankAmericard had already licensed hundreds of banks nationwide to issue its cards — but banks unwilling to join the Bank of America camp were shut out. Bank of America imposed extremely strict capital and scale thresholds on its licensing partners. Small banks didn't qualify for the "Visa club."
Those 16 locked-out banks, gathered in a conference room at Buffalo's Mark Hotel, founded an organization — the Interbank Card Association (ICA).
Their logic was brutally simple: we don't have Bank of America's national clearing network. But we can band together — and mutually accept each other's local credit cards. A card issued in Buffalo could be used in Pittsburgh. A card issued in Pittsburgh could be used in Kentucky.
ICA was the precursor to Mastercard.
The earliest product was called Master Charge. In 1979, it was renamed MasterCard.

The biggest problem this fledgling alliance faced wasn't Visa — it was trust. Every issuing bank was afraid the others would steal its customer list. So they designed a unique set of "neutral clearing rules" — ICA itself would not issue a single card. Its sole authority lay in setting technical standards and processing interbank clearing. Member banks were both customers and shareholders.
By the time Visa restructured itself into a similarly decentralized model under Dee Hock in the 1970s — Mastercard had already been operating under this exact model for nearly a decade.
This "members own the network, the network issues no cards" model proved extraordinarily robust — because it allowed every issuing bank to issue cards on both the Mastercard and Visa networks simultaneously. Most banks chose to issue both. That meant Mastercard didn't need to "poach" issuing partners from Visa — it only needed to provide a second channel that was mutually non-exclusive with Visa's.

In 1969, Mastercard did something nobody noticed but that was exceptionally clever — it switched its settlement and clearing system from nightly batch processing to a near-real-time online authorization system. It beat Visa's real-time authorization rollout by several years. Although Visa's network was larger and had broader merchant acceptance, Mastercard had a genuine bright spot in the early days when it came to moving faster on technology.
In the 1990s, Visa and Mastercard expanded into global markets in near-lockstep. But what truly transformed Mastercard from "the other card" into "the card some consumers actually prefer" — was the Priceless ad campaign.
1997. Mastercard's marketing team sat in a conference room at McCann Erickson, facing a wall of research. Consumers had told them, plain as day: every credit card does the same thing. Functionally, nobody cared about the difference between Mastercard and Visa.
So competing on features was pointless. They had to create an emotional connection.
The "Priceless" campaign said nothing about interest rates. Nothing about benefits. Nothing about credit limits. What it said was —
"There are some things money can't buy. For everything else, there's Mastercard."
That campaign ran for over 25 years and was translated into more than 50 languages. It recast Mastercard's identity from "a debit/credit card processing network" into "part of the experience of living."
The moment consumers started cracking their own jokes in the cadence of "Priceless" — Mastercard had already won.

Three things.
First: the survival rule for second place — never compete with number one on number one's terms.
Mastercard never tried to prove it was "more widely accepted than Visa." In 1997, it deliberately walked away from the features battle — and pivoted to the emotional battlefield. While Visa was still hammering "accepted everywhere," Mastercard stole the entire narrative around happiness with "Priceless."
What's the dimension you're competing on right now — and is it possible that it shouldn't be the same dimension your rival is fighting on?
Second: a coalition is harder to destroy than an empire.
Mastercard's membership structure means every issuing bank is both its customer and its owner. To destroy Mastercard, you'd have to persuade thousands of banks around the world to betray it simultaneously. That's basically impossible.
Is your business also building "collective immunity" — where every participant becomes a shareholder?
Third: second place doesn't need to outrun first place — it just needs to outlive everyone else.
From 1966 to 2025, the credit card network industry winnowed from thirty or forty players down to just two global giants. Mastercard didn't beat Visa — it won by outlasting every single one of the dozens of failed card organizations.
You're not fighting number one. You're sharing the podium with him. And there are only two spots on that podium.

Mastercard didn't invent the credit card. Didn't invent the clearing network. Didn't invent the member-owned alliance. It simply took everything that had already been invented — and executed it a little more elegantly, a little more enduringly, a little more "pricelessly" than the first guy. And 60 years later, when you scan the Global 500, you might find yourself running your finger down the table just to double-check: Mastercard's market cap — higher than Bank of America, higher than Citibank, higher than Goldman Sachs.
It's just a set of rules, a piece of plastic, and a tagline everybody knows by heart.

← Previous Contents Next →
Chapter 22 Caterpillar

In 1904, in a California marsh, a steam tractor sank into the mud, its wheels spinning uselessly in the slurry. The farmer crouched beside it, cursing. The photographer standing behind him, Benjamin Holt, didn't help him pull it out — he knelt down and stared at the wheel. There was only one question in his mind: what if a wheel weren't a wheel?

Not a new mechanical principle. Not a new energy source. Not a revolutionary breakthrough in materials science.
Just a photographer and farm-implement manufacturer who spent a few minutes staring at that mud that day, then went back to his factory and did one thing: removed the tractor's rear wheels. Replaced them with a looping, self-carrying "crawler track."
And then the machine could crawl across any mud, any sand, any hillside — like an endless steel caterpillar.
Holt watched it crawl and said: "That thing looks like a caterpillar."
That name has been painted on every Caterpillar bulldozer ever since. Caterpillar's market cap today: $413.6 billion.
Let me tell its story first. When I'm done, we'll talk about why the most important technological innovation — isn't making things more complex, but overturning the most basic mechanical intuition.

The day before Thanksgiving, 1904. The Sacramento–San Joaquin River Delta, California. Peat soil — the most fertile and also the softest soil on earth. Horses couldn't walk on it. Wheeled tractors sank into it. Only humans, standing on wide wooden planks, could tread on it without breaking through.
Benjamin Holt was 55 that year. He'd spent most of his life building agricultural steam engines — those giant iron-wheeled machines that crept across fields like locomotives. But the delta farmers were nearing despair — they had machines they couldn't use.
Holt drove a 40-horsepower steam tractor into the mud for a test. The rear wheels sank. The iron wheels spun uselessly in the slurry. He crouched beside it for over 40 minutes — then stood up, walked back to the factory, picked up a piece of chalk, and sketched a diagram on the blackboard: remove the rear wheels, replace them with a closed-loop chain fitted with gears at both ends — with wooden cross-planks laid across it. When the drive gear turned the chain, the track plates laid down their own "road" — a road surface that continuously unfolded ahead of the machine.
His chief engineer thought he'd lost his mind. "You want to build an endless road and mount it onto the vehicle?"
Holt didn't respond. He started modifying it directly.
1905. The first crawler-track tractor ran 8 miles across that peat soil without sinking once. Its ground pressure was only 6 pounds per square inch — less than the pressure of a human standing on the same ground.

Then the First World War broke out.
Europe's Western Front — stretching from Belgium to Switzerland — was a sea of mud churned up by millions of soldiers between the opposing trenches. Trucks, warhorses, human labor — all useless in the waist-deep Flanders clay.
Holt shipped his tractors to Europe. British and French soldiers rode atop the crawler tracks across shell craters. Then they mounted that tracked chassis inside a steel shell, added a gun — and the world's first tank was born. The name "tank" itself came from the British intelligence cover story: they wrote "Water Tank" on the shipping manifest to keep it secret.
So the tank, and the internal-combustion tracked-vehicle technology, came from a farm bulldozer conversion. Holt had no idea at the time that he had just invented the central machine of 20th-century ground warfare.

1925. Holt's Holt Manufacturing Company and his fiercest competitor, C.L. Best's tractor company, merged — forming the Caterpillar Tractor Company. These two men had been suing each other over track patents for over a decade. But in a moment of post-war agricultural collapse, they chose to point their guns outward together — and take on the global market as a united front.
From that point on, Caterpillar became synonymous with "tracked vehicles." The Great Depression, the Marshall Plan, post-war reconstruction, globalization — every time civilization rebuilt itself, yellow Caterpillar bulldozers were there, leveling foundations in the mud.

Now stop here for a moment.
There's a question you've definitely encountered in your own industry — is there some fundamental assumption, some "that's just the way it is" given — wheels are round — that no one has ever questioned, until one day you crouch down and look at it and think: what if it isn't a wheel at all?
Holt didn't invent the mathematics of the continuous track. He just crouched in the mud for 40 extra minutes.

Three things.
First: it wasn't invention — it was "modifying existing components to the extreme."
Holt invented neither the steam engine nor the internal combustion engine. All he did was add a chain-belt that laid down its own road to the undercarriage. He wasn't an original theorist — he was an integrator who connected two known components in a way nobody else had.
How many known components are sitting in your hands right now — waiting to be paired in ways no one has tried?
Second: turn your competitor into a merger partner.
Before merging, Holt and Best had been litigating against each other for years. After merging — both sides shared all of their patents and distribution networks. A fight to the death became unified global pricing power.
Is there any external threat right now, between you and your main rival, large enough to make it worth sitting down and discussing a united front?
Third: name your most fundamental component after your brand.
"Caterpillar" is no longer an insect. It's a moving ground. When bulldozers on job sites all over the world are simply called "Cats," it's no longer a product. It's a species.
Has your brand — already become that verb?

Today, the track on every Caterpillar bulldozer still operates on that 1904 principle: the machine carries its own road with it, a road that continuously unfolds beneath its body.
It doesn't build machines. It builds the first road in places that have no roads.

← Previous Contents Next →
Chapter 23 Cisco

On the Stanford campus, a married couple built a router in their lab — because different departments couldn't email each other across buildings. Their department chair said it had no commercial value. Fifteen years later, Cisco was the most valuable company on earth.

Not market research. Not a technological breakthrough. Not capital engineering. Not a business plan.
Just Sandy Lerner and Leonard Bosack. Two Stanford IT administrators. In 1984, they strung the first multi-protocol router between the business school and the computer science department. They just wanted to transfer files to each other — because Sandy was in a business school building, Len was in a computer science building, and the two buildings' networks didn't speak each other's language.
Then they started selling the thing out of their garage. Within 15 years, it became the most important internet infrastructure company on the planet. Cisco's market cap today: $388 billion.
Let me tell their story first. When I'm done, we'll talk about why the most important inventions often come from "I just wanted to solve my own little headache."

1984. The Stanford University campus.
Sandy Lerner was the director of computing for the business school. Leonard Bosack was the network administrator for the computer science department. They met through work. The two of them constantly needed to transfer data between their respective buildings — but they couldn't. Because the business school used the DEC network protocol, while the computer science department used TCP/IP. (Yes, in 1984, TCP/IP was not yet the standard.)
The Stanford administration said: you cannot run your own cable.
They said: we didn't run cable. We wrote a cross-protocol translation program on the router.
They weren't building a commercial product. They were solving their own frustration.
They soldered circuit boards in their living room. The router motherboard came from an electronics store. The operating system and the protocol translation were all written by hand.
And then they discovered — it wasn't just the two of them. Every department at Stanford, every competing computer faction — IBM's SNA, DEC's DECnet, Xerox's XNS, and TCP/IP, still struggling on the margins — all jammed against each other by language barriers.
1986. They built the AGS — the first multi-protocol router. Not one protocol to another. All protocols translating with each other simultaneously.
Then they moved Cisco (the name comes from the tail end of "San Francisco" — "cisco," all lowercase) into a garage in Menlo Park. The two of them maxed out credit cards to buy components. Sandy handled sales and customers. Len handled production.

Their first year, they picked up an unexpected customer — universities. Not just Stanford. Universities, defense contractors, federal labs all over the world — every single one had the same problem: computers from multiple manufacturers couldn't interconnect. There was no commercial product on the market that could do multi-protocol routing. Cisco was the only one.
Then Boeing called. Then General Electric. Then Goldman Sachs.
Then came the 1990 IPO — on listing day, Cisco was valued at $224 million.
And then — Sandy and Len were fired.

August 1990. Cisco had been public for only six months. The venture-capital-backed board — led by Don Valentine of Sequoia Capital — told Sandy and Len: you are not the people to run a $10 billion company. We need a professional CEO.
The couple fought back fiercely. Not for their own positions — but for the engineer-driven culture they had created. They believed that product decisions should be made by "those with the deepest technical intelligence," not by budget processes dictated by a P&L statement.
Don Valentine's view was the polar opposite — "Genius engineers can create products, but they cannot control an institution that is already reporting to Wall Street." He brought in John Morgridge as CEO.
After Morgridge arrived, Sandy submitted her resignation. A few weeks later, Len followed her out the door. They sold virtually all of their stock around 1991 — had they held until 2000, they might have been one of the wealthiest couples in computing history.
The day Sandy left, Morgridge said a line that Cisco people have repeated for thirty years: "Sandy and Len gave the world the internet's translator. We just helped them wrap it up and ship it to every airport and train station."

Now, the question is: if the two founders hadn't left — would Cisco still have become Cisco? After Sandy and Len departed, Cisco entered a frenzied era of M&A-driven growth — buying more than 70 companies between 1993 and 2000. A company that bought companies. It swallowed every technology adjacent to "routing" — switching, voice, security, wireless — shoveling them all into its belly with checks. By March 2000, Cisco was the most valuable company on earth — $569 billion.
Then the dot-com bubble burst. Cisco plummeted from $569 billion to under $100 billion. But it got back on its feet amid the rubble — and didn't die. Because the technology from all those companies it had acquired had already become the nerve endings of the internet.

Now stop here for a moment.
If you had to step back once your company reached a certain size — handing your creation over to a professional CEO — and then watch it balloon to heights you could never have reached on your own — would you step aside?
Sandy and Len's story doesn't end in romance. But they gave the world the most important multi-protocol router ever built. And Don Valentine and his professional CEO carried it into every corporate building on earth.

Three things.
First: solve your own problem — and then discover the whole world hurts the same way you do.
Sandy and Len didn't do "market research." They were simply driven crazy by the chaos of the Stanford campus network. In your daily life — is there something jamming you, hurting you every single day — that you've grown so used to, you no longer think of it as something that should be a product?
Second: building a "translator" pays better than building a "new language."
Cisco didn't invent any network protocol. It built a translator that let all protocols speak to each other. When an industry has multiple standards that can't interoperate — don't pick a side. Build the bridge. Bridges collect tolls.
Where is the "multi-protocol scrum" in your industry? Can you be that bridge?
Third: a router is not a product — it's the neural node of the internet.
The key to Cisco's rise in the 1990s wasn't selling more boxes. It was selling "the machine every single email must pass through." It wasn't a product company. It was an infrastructure company — once installed in a server room, no ops team dared touch it — "touch a Cisco box, and the whole network goes down."
Once your product is deployed, are your customers too scared to unplug it either?

Sandy Lerner eventually came back — in 2023 she appeared in the news with her new company and her passion for organic farming. Her life has never been defined by anyone else's terms. And that little blinking blue light from Stanford is still routing every data packet in the world.

← Previous Contents Next →
Chapter 24 Netflix

In 1997, he rented Apollo 13 and returned it six weeks late. The late fee was $40. He walked out of the video store with a single thought in his mind: I'm going to kill this industry.

Not disruptive technology. Not the internet revolution. Not the invention of streaming. Not big-data algorithms.
Just Reed Hastings. A man enraged by a $40 late fee. He walked out of Blockbuster, tossed the tape onto the passenger seat, and spent the entire drive home thinking — how do gyms charge? A monthly fee. Pay once, go as many times as you want. Why couldn't video rentals work the same way?
And then, with that idea sparked by a $40 fine, he put Blockbuster — the world's largest video rental chain, 85,000 employees, 9,000 stores, a $6 billion market cap — into the grave.
Then he personally incinerated his own most profitable business — pivoting from DVD-by-mail to streaming — and then blew up the entire Hollywood distribution model.
Netflix's market cap today: $374 billion.
Let me tell its story first. When I'm done, we'll talk about why "killing yourself" is the hardest muscle for a founder to build — and one you have to flex every few years.

1997. Santa Cruz, California.
Reed Hastings was a mathematician and software engineer. His first company, Pure Software, had sold to Rational Software in 1997 for $700 million. He made money, but he believed he'd been a terrible CEO — he didn't understand management.
He reflected for a long time and distilled one lesson: the best teams don't need to be managed — what they need is density. Pack A-players together. The moment a B-player walks in, everything breaks down.
This later became the original prototype for Netflix's famous "talent density" culture.
But that day, his mind was not on management philosophy. It was on $40.
He sat on the floor of his living room and stuffed a few music CDs into envelopes — and mailed them to a friend. A few days later, the CDs arrived intact. USPS First Class mail. You could mail a CD with a single stamp.
He thought — isn't a DVD exactly the same size?
He and Marc Randolph tested the idea together. They slipped a DVD of My Fellow Americans into a regular envelope and dropped it in the mail. A few days later, the DVD arrived at the other address, undamaged.
Then they built a website with only two operations: pick a movie, return a movie. No late fees. Monthly subscription.
Consumers didn't need to go to a store. Didn't need to return anything. Didn't need to awkwardly explain to the clerk why it was late. (At the time, late fees accounted for 16% of Blockbuster's revenue.)
In 1999, they tried to sell the company to Blockbuster for $50 million. Blockbuster CEO John Antioco listened politely. Then he smiled and said no — reportedly, as he walked out of the meeting, he told his executives, "It's nothing more than a niche mail-order business."
2002. Netflix went public. 2004. Blockbuster finally realized it had been wrong and pivoted hard into online DVD-by-mail — but it was already five years too late.
2010. Blockbuster filed for bankruptcy protection. That year, Netflix surpassed 20 million subscribers.

But the wildest pivot came after that.
2011. Hastings did an internal analysis. The bandwidth cost of streaming was plummeting, while the postage and warehousing costs of DVD-by-mail remained flat. He drew a crossover line.
He announced: Netflix would be split into two products — the DVD-by-mail service would be called Qwikster. The streaming service would be called Netflix.
The user reaction was explosive. Angry emails. A world war on social media. Front-page mockery from every newspaper. Netflix lost 800,000 subscribers in a single week. The stock price dropped 77%.
A company hailed from 2008 to 2010 as one of the strongest businesses in history, within the space of weeks, looked like a ship steering itself into an iceberg.
In his apology video, Hastings didn't defend himself. He only said one thing: "We moved too fast. But we haven't changed direction — we just stumbled on the footwork."
Then he kept turning. Faded out the DVD business entirely. Moved streaming bandwidth costs onto AWS. Poured every remaining dollar into original content.
2013. House of Cards went live.
That was not a TV show. It was a nuclear bomb. A political thriller. Directed by David Fincher. Starring Kevin Spacey. Netflix spent $100 million on the first season. Not $1 million. Not $10 million. One hundred million dollars.
Every Hollywood player, at every Hollywood party, laughed and said — these Silicon Valley tech guys are pouring venture capital money down the drains of the Chateau Marmont.
Three months after House of Cards aired, no one was laughing. Netflix added 3.3 million global subscribers. Overnight, it went from being HBO's distributor to being HBO's competitor.
Then Stranger Things, The Crown, The Queen's Gambit, Wednesday... hit after hit. It became one of the largest content factories on the planet, spending over $17 billion a year on original content. The six major Hollywood studios — Disney, Warner Bros., Universal, Paramount, Sony, Fox — suddenly discovered a Silicon Valley stranger standing on their turf.

Now stop here for a moment.
Can you, while your most profitable business is still perfectly healthy — draw that crossover line, calculate when it will begin its death spiral — and preemptively swap it out for the next model?
Hastings began shifting to streaming while DVD-by-mail was still growing. He began investing in original content while streaming licensed content was still profitable. He started the self-revolution while he was still in the lead.
Your current core product — have you ever drawn its "death countdown clock"?

Three things.
First: the customer doesn't want a DVD. The customer wants a movie.
Too many founders mistake the physical form of their product for what the customer actually needs. Blockbuster thought it was selling "a VHS tape inside a rental store." Netflix knew it was selling "you can watch any movie without leaving your house."
The specific thing you sell — what is it, really? Strip away the physical form: what does the customer actually want?
Second: talent density.
In the Netflix culture deck — the one Facebook's Sheryl Sandberg called "possibly the most important document to come out of Silicon Valley" — Hastings wrote one line: "We run the company as a professional sports team, not a family. If you're not a star — we'll give you a very generous severance package and give your spot to an A-player."
In your team, how many people are there whom you "can't bear to lose" rather than "absolutely must have"?
Third: making originals doesn't mean "making it yourself" — it means "swallowing the most expensive link in your supply chain."
At first, Netflix simply bought streaming rights to other studios' content — it was essentially a Hollywood supermarket with distribution rights. Then licensing fees kept climbing, and Hollywood began pulling its own IP back onto its own streaming platforms. Hastings realized: if he didn't make originals — his shelves would be stripped bare.
So he went from being a "middleman" to being "the source of content." Not because he was a master storyteller — but because his supply-chain analysis told him that without moving upstream, there would be no future profit.
In your supply chain — what's the most expensive link? Can you do it yourself?

Reed Hastings once said: "Our biggest competitor isn't HBO, isn't Disney+ — it's our users' hours of sleep."
He's not in the video business. He's in the war for human attention. And the total supply of attention is fixed.
What human resource are you truly competing for?

← Previous Contents Next →
Chapter 25 Chevron

In 1911, the U.S. government took a knife and carved Standard Oil into 34 pieces. One of those pieces was Standard Oil of California — today, it is called Chevron, with a market cap of $370 billion.

Not a startup built from scratch. Not a garage venture. Not a technological invention. Not a business-model innovation.
Just a single antitrust ruling. The U.S. Supreme Court said Standard Oil was too big and had to be broken up. And in a single day, Rockefeller's giant empire was sliced into 34 independent companies. Each piece carried away a piece of the parent's DNA — extreme cost control, vertical integration, and an obsession with efficiency.
More than a hundred years later, among those 34 fragments, ExxonMobil and Chevron have grown into giants even larger than the original Standard Oil.
This story is not about starting a company. It is about a corporate DNA that, after being dismembered, not only refused to die — but reproduced independent life in every single fragment.
Of all the companies ever broken up — this is the most successful case.

Standard Oil of New Jersey → Exxon. Standard Oil of New York → Mobil. Standard Oil of California → Chevron. Standard Oil of Indiana → Amoco.
After the breakup, each inherited one of Rockefeller's operating principles: own everything, from the oil field to the gas station — "no leasing, no outsourcing, no one gets to choke your throat." Vertical integration.
Chevron's particularity was that from the very beginning, it tilted upstream — exploration and extraction. California, Texas, Alaska, and then Saudi Arabia.
In 1933, Standard Oil of California acquired the oil exploration rights to Saudi Arabia. This was the first substantive extraction contract an American oil company ever signed in the Middle East. Saudi Arabia would later become the kingdom with the largest proved oil reserves on the planet.
In the 1940s, the California-Arabian Standard Oil Company (controlled by Chevron) discovered industrial-grade oil at the Dammam field — and was then forced to share its Saudi stake with Texaco, Exxon, and Mobil (because the U.S. government needed collective power to stabilize the post-war Middle East). This was the origin of what later became Aramco — Saudi Aramco.
Today, Saudi Aramco is the most valuable company on earth — and Chevron was once, in a legal sense, its "father."
Over the 80 years that followed, Chevron integrated relentlessly: it acquired Gulf Oil (1984, then the largest merger in history), merged with Texaco (2001), and then kept drilling, well by well, across South America, Australia, and Kazakhstan.

But the Rockefeller tradition Chevron truly inherited wasn't exploration rights — it was cost paranoia.
One thing many energy analysts have noted: among the six oil supermajors, Chevron's unit production cost has consistently been among the lowest. Its capital discipline resembles Rockefeller's desk — forever calculating the per-barrel unit cost. Never indulgent.
That's why, when crude oil prices collapsed in 2020, Chevron's balance sheet didn't blow up — while many competitors were forced to cut dividends and lay off workers.
Driving costs down — what Rockefeller was doing in Cleveland 150 years ago — Chevron is still doing today in the shale well sites of the Permian Basin.

Three things.
First: corporate DNA can reproduce after dismemberment.
Every single fragment of Standard Oil turned into the same kind of "cost control + vertical integration" company. What Rockefeller created was not a company — it was an irreversible organizational culture. When the law shattered it to pieces, the culture replicated itself.
If your company were broken into ten pieces tomorrow — could your corporate culture survive independently in every one of them?
Second: keep your eyes fixed on the upstream cost structure.
Chevron's money wasn't made at the gas pump. It was saved, fifty cents at a time, from the oil fields and the refineries. Its capital discipline over the past decade has given it more dry powder to buy cheap assets in the next industry downturn.
Across your entire value chain — at the link closest to "raw materials," is your cost structure the lowest in the industry?
Third: political risk can't be managed by a compliance department — it has to be hedged with "full-chain control."
Oil is one of the most politically sensitive industries in the world. Every time Chevron enters a country, it doesn't just sign a licensing deal and walk away. It bundles together technology, pipelines, refining, and retail terminals — so that if a local government defaults, the fastest reaction doesn't come from a courtroom; it comes from gas stations running dry.
In your international markets — have you kept the power of supply inside your own body?

Chevron has no founder story. It has an ancestor story. And its ancestor — was hacked into 34 pieces by the law — and every single piece turned into an immortal species.

← Previous Contents Next →
Chapter 26 AbbVie

Some call him the world's most profitable "patent thief." But every company he stole from ended up making even more money.

Not inventing new drugs. Not breakthrough R&D. Not a scientific genius. Not an M.D.
Just Richard Gonzalez. A Cuban-American biochemist. He had a very simple creed — inside the labs of big pharma, there are countless half-abandoned molecules. Killed because the marketing department decided "the market isn't big enough." He picked up these discarded molecules and turned them, through AbbVie, into super-blockbusters.
Humira — the single highest-selling drug in human history — came about exactly this way. It wasn't invented by AbbVie. It was a half-finished molecule that Abbott bought from BASF. And then Gonzalez turned it into the "king of drugs," generating $20 billion a year in sales.
In 2013, AbbVie was spun off from Abbott and listed independently — today its market cap is $370 billion.
Let me tell this different story first. It has nothing to do with garage startups. It's about — how to do something inside an already-big company that "the big company says can't be done."

The 1980s. The labs of Abbott Laboratories.
A few German chemists in a BASF lab had synthesized a new human monoclonal antibody — D2E7. It targeted an inflammatory factor called TNF-alpha. In plain English: the reason rheumatoid arthritis patients' joints become inflamed, painful, and deformed is that the concentration of TNF-alpha in their bodies is too high.
BASF pushed the molecule through Phase I clinical trials, then decided the commercial prospects weren't big enough — the arthritis market genuinely wasn't large at the time. So they sold the molecule cheap to Abbott.
Even inside Abbott, not everyone was bullish. An arthritis drug? The ceiling on that market is tiny. And it requires long-term injections — will patients accept that?
Richard Gonzalez — then head of pharmaceutical R&D at Abbott — insisted on pushing forward. He used a very simple argument:
"TNF-alpha is not just about arthritis. It's a core pathogenic factor in Crohn's disease, psoriasis, ankylosing spondylitis. We're not developing an arthritis drug. We're developing a 'universal key for autoimmune diseases.'"
Humira was approved in 2002. Then Gonzalez's team did one thing — over the next 15 years, they systematically applied for and won approval for Humira across virtually every autoimmune disease indication. Rheumatoid arthritis. Psoriatic arthritis. Ankylosing spondylitis. Crohn's disease. Ulcerative colitis. Hidradenitis suppurativa. Uveitis...
With each new indication, Humira's market doubled.
And with each new indication, Humira's period of patent exclusivity in that indication could begin counting anew. This was the "patent thicket" strategy — not relying on a single patent, but densely planting hundreds of patents around the drug molecule covering manufacturing processes, formulations, and indications — so that generic manufacturers couldn't find a single crack to slip through.
By 2021, Humira's cumulative sales had surpassed $200 billion. No drug in human history has ever sold more.

In 2011, Abbott CEO Miles White did something that had every Wall Street analyst shaking their head.
He split Abbott into two companies — one would continue selling baby formula, diagnostic equipment, and generics (keeping the Abbott name), and the other would focus on innovative patented drugs (named AbbVie).
The logic behind the split: investors won't give innovative-pharma valuations to a mixed company that sells "baby formula + patented drugs." They'd compress the P/E ratio down to consumer-goods levels.
After the split, AbbVie immediately commanded a pure innovative-pharma valuation upon listing — its P/E ratio was several times higher than before.
And that D2E7 molecule bought from BASF became AbbVie's biggest cash engine in its first decade. Then AbbVie used Humira's cash to buy Pharmacyclics (maker of the blood cancer drug ibrutinib) and Allergan (maker of Botox) — expanding its pipeline from immunology into oncology and medical aesthetics.
When Humira lost its U.S. market exclusivity in 2023, analysts once again predicted AbbVie's profits would fall off a cliff. Then Skyrizi and Rinvoq — two next-generation immunology drugs already bet on during the Gonzalez era — filled most of the gap.
The cycle ran again: old drug expires, new drug steps up, then keep acquiring, keep building patent thickets.

Now stop here for a moment.
Do you have your own "Humira" — something abandoned by a peer or a predecessor, something that looked too small or too troublesome — but if you took it to the extreme, it might become your longest snow-covered slope?
How many projects have you abandoned because they "didn't look good enough to sell" — only because their starting point wasn't big enough?

Three things.
First: pick up the molecules others have thrown away.
BASF gave up on D2E7. Half the people inside Abbott wanted to give up on it too. But Gonzalez took it. He wasn't betting on the molecule — he was betting that the TNF-alpha target would link dozens of diseases together.
In your industry, is there something "everyone else thinks has no future, but the target is right"?
Second: a patent thicket isn't a legal question — it's a strategy question.
Humira isn't one patent. It's over a hundred. Add one for each new indication. Add one for each manufacturing step. Add one for each minor formulation tweak. This is Go, not chess. What you're surrounding is time.
Can your patent/moat portfolio surround a 15-year window of exclusivity?
Third: sometimes a breakup creates value.
Abbott voluntarily spun off AbbVie — not because the business was struggling, but because "being mixed together depresses the valuation." After the split, both companies multiplied in value. Inside your current business, is there a high-growth division — being suffocated under the valuation weight of a conservative division?

Richard Gonzalez later said: "Truly great drugs are usually not invented — they are discovered. And the person who discovers them is usually the one willing to look into the petri dish one more time after everyone else has already walked away."

← Previous Contents Next →
Chapter 27 Bank-of-America

The 1906 San Francisco earthquake. The entire city was burning. Every bank was scrambling to haul its gold bars out of town. Only one short Italian immigrant — charged in the opposite direction, straight into the ruins. He stuffed $80,000 in gold bars into a vegetable cart, covered them with a layer of oranges, and drove through the burning streets back to his tiny shop. The next morning, he set up a wooden table on the wharf with a sign — "Need money? Give me your hand to sign."

Not capital engineering. Not mergers and acquisitions. Not government connections.
Just A.P. Giannini. An Italian immigrant who grew up on the San Francisco wharves. He spent his entire life doing something every big bank looked down on — lending money to "the little guy." To dockworkers, vegetable farmers, laundry workers, small shopkeepers. Immigrants who'd been in America for three years and still couldn't speak fluent English — in the eyes of the big banks, they weren't even qualified to open an account.
And then he gathered up all that "small change not worth serving" and built Bank of America. Today, Bank of America has a market cap of $356 billion and is one of the largest consumer banks in the United States.
Let me tell his story first. When I'm done, we'll talk about why "looking down on" are the two most expensive words in the business world.

Amadeo Pietro Giannini. Born in 1870 in San Jose. When he was seven, his Italian immigrant father was shot dead by a farm worker in a dispute over a $1 gambling debt. His widowed mother later remarried a produce hauler. The young Giannini woke up every morning at 2 a.m., hitched up a wagon, and drove to farms to collect fruits and vegetables — then hauled them to the San Francisco wharf to sell to wholesalers.
He spent his whole life watching who actually worked. Who really got up before dawn and came home after dark.
He saw it clearly: the banks wouldn't lend to these people — not because they couldn't repay. Because the banks couldn't be bothered to verify their income. A dockworker might only earn a few hundred dollars a year — but that was a steady paycheck, arriving every two weeks. There just wasn't a "credit report."
In 1904, he opened a small shop on Washington Street in San Francisco's North Beach. The sign read: Banca d'Italia — the Bank of Italy. Hours: 7 a.m. to 7 p.m. The big banks opened at 9 a.m. and closed at 3 p.m. — exactly overlapping with working people's shifts. He kept the doors open in the evening — so workers could deposit money after their shifts.
He didn't work out of a marble lobby. He knocked on doors, one by one: "Sir, you earned $2 today. Keep $1 to feed your family — give me the other dollar. I'll save it for you. In a year, you can buy a steamship ticket to bring your family over from Italy."

Then came April 18, 1906. 5:12 a.m. The great San Francisco earthquake. Followed by four days and four nights of fire that burned 80% of the city. Every big bank was racing to move its vault — but the fire spread too fast, and many couldn't even get to their vaults in time.
Giannini rolled out of bed and jumped to his feet. He commandeered a vegetable cart from San Mateo and charged into the still-burning financial district. The building housing his bank was already smoking — he stuffed $80,000 in gold bars, silver coins, and ledgers into burlap sacks, covered them with oranges from the fruit crates, threw a canvas tarp over everything, and drove through the fire cordon.
By 7 a.m. the next morning, he had set up a wooden table on the Washington Street Wharf in North Beach. Beside it stood a wooden sign he'd lettered himself in charcoal: "Need money?"
All the big banks were closed — because their records had burned, their vaults were destroyed, their halls had collapsed. And the big banks' charters stipulated: no loan without full collateral.
Giannini had no charter. He issued loans on the spot. The only collateral — a handshake. He told every person who came to borrow: "You rebuild. I'll take your hand as collateral."
Over the next three months, he issued more than 2,000 reconstruction loans. Every single one was repaid in full within a year — most of them early.
The citizens of San Francisco discovered, for the first time, that "high-risk" poor people repay their loans more punctually than "low-risk" rich people and institutions.
From that day forward, Giannini's multi-branch model expanded at a furious pace. He opened branches in every California small town where someone grew vegetables or ran a shop. Every branch manager had to speak the language of the local immigrant community. Cantonese in the Chinatown to the north. Italian in the vineyard country. Spanish in the Central Valley.
In 1928, sensing the Great Depression was coming, he converted his entire stock portfolio into hard assets. When the market crashed in 1929, his vaults were filled with real gold and bonds. Then, through the 1930s, he bought up every competitor that went bankrupt in the collapse — at rock-bottom prices.
By the time he died in 1949, Bank of America was the largest private bank in the world.

Now stop for a moment.
Among your customers, is there a group "the big players won't touch"? If you did every single transaction that others consider "beneath their standards" — could you build a torrential foundation in the grassroots that no one else cares about?
The big banks have always been competing with each other for "the wealthiest 1% of customers." Giannini never fought on that battlefield for a single second. He took the other 100%.

Three things.
First: don't look at the collateral — look at the person.
Giannini's only risk-control metric was: what does this person's neighbor write about them in a letter? Branch managers had to live locally, their kids had to attend the local schools — they had to eat dinner on that street at least twice a week. If the neighborhood said this person was good — they were good.
Is your current risk management — too reliant on numbers, and not enough on the judgment of people who "live on the same street"?
Second: branches are not cost centers — they are community stabilizers.
Giannini's logic for opening branches was peculiar — he didn't pursue profit maximization. He pursued density. So that an immigrant worker didn't have to take a ferry across town just to deposit 50 cents.
The key insight: when customers didn't have to leave their neighborhood to deposit money — they deposited. Once they deposited, the bank had a deposit base. Deposit base = cheap money = ammunition to expand lending.
Are your touchpoints or outlets — dense enough that customers find "taking one extra step" too much of a hassle?
Third: the one who releases water during a panic — is the biggest winner when the panic is over.
The 1906 fire, Giannini lent. The 1929 crash, Giannini swept up assets. Every panic, everyone was selling — and one Italian immigrant was hauling gold in a cart of oranges.
Are you ready? When the next panic hits your industry — will you be that "orange cart"?

Giannini died in 1949. No banker gave a eulogy at his funeral. Dockworkers and vegetable farmers filled the street from end to end. Someone carved a line into his tombstone: "He lent money to those who couldn't borrow it anywhere else." He had no heirs. He donated his entire stake to a foundation.

← Previous Contents Next →
Chapter 28 Lam-Research

In 1980, Silicon Valley. An engineer who had spent over a decade etching chips at Fairchild and Intel — David Lam — sat at his kitchen table and told his wife: "I'm quitting. I'm going to build something everyone says nobody needs — a machine that does nothing but etching." At the time, the entire semiconductor industry was selling "all-in-one equipment." A single machine had to deposit films, etch, clean, inspect — because fabs wanted "one machine per room that does everything." Lam said: "Wrong. Etching should be broken out and pushed to the limit." Today, Lam Research has a market cap of $350 billion and is one of the absolute monopolists in global plasma etching equipment.

Not a supply-chain shift. Not the semiconductor cycle. Not TSMC investment. Not China's chip strategy.
Just David Lam. An engineer who stood in Intel's cleanroom watching plasma etch silicon wafers for ten years. He saw something every equipment maker saw but no one acted on — that as line widths shrank, etching was becoming a game of physical limits, and "all-in-one machines" were getting more mediocre by the day. And then he made that classic Silicon Valley move — quit, mortgaged his house, and built a machine that "only does one thing."

Lam was born in Taiwan in 1945 and later came to California State University to study engineering. When he joined Fairchild, that company was Silicon Valley's "transistor engineering academy" — its alumni would go on to found Intel, AMD, and dozens of semiconductor equipment companies. Lam was responsible for the etching process in the cleanroom — the step where plasma "carves" the pattern from the photoresist into the surface layer of the silicon wafer. In the 1970s, this job was basically manual — an engineer stood next to the machine and adjusted gas flow rates by hand, twisting valves. Lam stood in front of that machine every day tuning parameters, and gradually began to develop more curiosity about the machine itself than about people.

The crack he found was this: etching was shifting from "wet chemical etching" to "dry plasma etching." And all the major equipment makers — Applied Materials, Perkin-Elmer — were building "integrated equipment," because they believed chip fabs wanted one-stop shopping. Lam believed the opposite — as line widths shrank below one micron, etch uniformity, selectivity, and profile control would become so fiendishly difficult that a company that "only does etching" could concentrate all its R&D resources on this single step and then devour the all-in-one competitors through sheer technical depth.

In 1980, he founded Lam Research. The first year, there were six employees — including his wife handling the books. He named his first plasma etching machine the AutoEtch — its core selling point wasn't "cheaper" but "the edges of the etched lines were so clean, no one had ever seen anything like it." 1984, Lam went public. Over the next four decades, Lam kept pushing etching to the atomic level — from microns to nanometers, from planar to the deep-hole etching of 3D NAND. In today's advanced logic chip and 3D flash manufacturing, Lam's etching equipment forms a triumvirate with Tokyo Electron and Applied Materials — and Lam, to this day, still "primarily only does etching."

Three things.
First: take one thing that everyone bundles together — break it out and do it to an insane level of perfection.
The founding principle of Lam Research is brutally simple — semiconductor manufacturing has hundreds of steps. He picked one. And only did that one. And that one became so large in volume that it could independently support a $350 billion market-cap company.
In your industry — is there a step currently "bundled" into a generic solution that's worth breaking out into its own company?
Second: stand in the cleanroom and watch — don't sit in your office listening to reports.
Lam's understanding of etching didn't come from market reports — it came from standing in front of the machine for ten years. He smelled the plasma. He touched the deposits inside the etch chamber. He heard the abnormal vibrations of the vacuum pump.
When was the last time you "stood on the production line yourself"?
Third: betting small — is a strategy, not a retreat.
Lam not building "all-in-one equipment" looked like "giving up on the big market." In reality, the share he ate up in the single category of plasma etching is higher than the total revenue of many "broad but shallow" equipment makers. Narrowing your scope — if that scope happens to be the industry bottleneck — can earn far more than broadening it.

Lam is now retired. He invested in a handful of clean-energy startups in Silicon Valley — using plasma technology to produce hydrogen. He says etching and hydrogen production are fundamentally the same physical process — "you just swap the silicon wafer for a water molecule."

← Previous Contents Next →
Chapter 29 UnitedHealth

1974, Minnesota. A group of doctors were arguing in a conference room in suburban Minneapolis. The core of the debate was a sentence that sounded heretical at the time: "What if we let doctors — not insurance companies — decide what treatment is 'necessary'?" The man who spoke those words was Richard Burke. He was not a doctor. He was someone who had managed finances inside a health maintenance organization (HMO). He had just left his previous HMO — because he discovered that an insurance company could overturn three years of a doctor's clinical judgment with a single claim-denial form. And then he founded UnitedHealth. Today, UnitedHealth Group has a market cap of $349 billion and is the largest health-insurance-and-healthcare-services hybrid on earth.

Not government healthcare. Not an aging population. Not medical inflation. Not regulatory moats around insurance licenses.
Just Richard Burke. A man who saw that the "insurance denial letter" reached the patient faster than the "critical condition notice." And then he proposed a counterintuitive business model: if you let doctors manage the healthcare budget — instead of actuaries — you will spend less money over the long run, because good prevention is cheaper than cheap claims processing.

Burke was born in 1943 in Ohio and studied economics at the University of Minnesota. In the early 1970s, he worked at a nonprofit HMO — part of a wave of experimental health insurance organizations spurred by the Health Maintenance Organization Act of the Nixon era. He observed that the HMO's finance department managed doctors as a "cost center" — setting annual physician budget caps and penalizing those who exceeded them. Doctors, in turn, learned a set of counter-tactics: over-order tests but cut back on high-risk treatments. Both sides gamed each other, and the patient was caught in the middle.

Burke said: "Let doctors manage their own budgets — give them data, give them risk, and give them reward." This model later came to be called "physician-led care management" — UnitedHealth's core differentiator in its early days. He first bought a few small HMOs on the West Coast, then brought doctors onto governance committees — giving doctors access to "panoramic data on the medical costs of the entire patient population" and letting doctors themselves design prevention and chronic-disease management pathways to reduce long-term total costs.

This model was extraordinarily slow. When UHC went public in 1984, it had only a few hundred thousand members — while Kaiser Permanente at the time had several million. But Burke's patience lay in this: he spent ten years collecting hard data proving that "prevention saves money compared to treatment." Every case of early hypertension control, every diabetes dietary-management intervention, every nurse follow-up phone call — he filed away cost-comparison data on all of it. By the time the HMO market exploded in the 1990s, UnitedHealth possessed the industry's richest dataset on "cost seen from the doctor's perspective" and devoured the largest share of the incremental growth.

And then — UnitedHealth's Optum division became the real engine. Optum integrated medical data analytics, pharmacy benefits management, and physician-group operations into a single system — creating an apparatus that "intervenes in medical decisions before the insurance claim is even filed." This system today covers over one hundred million Americans — spanning the full chain from insurance to pharmaceuticals to clinics.

Three things.
First: hand the power of "controlling costs" — to the people who generate those costs.
Burke let doctors manage the budget. This wasn't "trust" — it was calculation. Because once a doctor sees the aggregate medical-cost data for his own patient population, he naturally develops the same cost-consciousness as a finance person — but with clinical judgment attached.
Is your profit control — in the hands of those who make the profit, or those who "manage" the profit?
Second: patiently collect counterintuitive evidence — and then wait a decade.
To prove that "prevention saves money," Burke stored ten years of data and medical records. Most people would have given up by year three — because short-term costs actually go up. He knew that "prevention saves money" is mathematically certain to be front-loaded on cost with a tail-end payback — so he prepared in advance the patience to explain that J-curve to investors.
Is there something you know is mathematically correct but whose short-term data looks unfavorable — waiting for you to "stockpile enough evidence"?
Third: enter before the claim is filed — not intercept after.
The core of Optum isn't "managing insurance" — it's using data and physician networks to influence decisions before the medical action even occurs. Can your business — enter the customer's decision before "the cost is incurred," rather than auditing after "the cost has already been spent"?

Burke stepped down as CEO in 2006. He later donated funds to establish a school of public health at the University of Minnesota — named the "Burke Institute for Public Health Leadership." Outside his office door hung a single line — "Health is not produced by claims processing."

← Previous Contents Next →
Chapter 30 Coca-Cola

1886, Atlanta. A pharmacist boiled a pot of brown syrup in his basement using a copper kettle. He sold it as medicine — for headaches, for fatigue, for neurasthenia. The first year, he sold only 25 gallons. But then someone came along and turned that pot of syrup into the most famous taste on earth.

Not inventing a formula. Not pouring money into advertising. Not channel expansion. Not capital engineering.
Just Asa Candler. A Georgia pharmacist and real-estate businessman. He suffered from chronic headaches. In 1888, he tried that pot of brown sugar-water concocted by another pharmacist, John Pemberton — and said his headache went away. Then he bet his entire fortune on that copper kettle, spending $2,300 to buy the formula.
And then he spent 30 years turning a local "brain tonic" into America's national drink. Today, Coca-Cola has a market cap of $341.8 billion, and the world consumes more than 2 billion servings every single day.
Let me tell this story first. When I'm done, we'll talk about why Candler's most brilliant move — wasn't selling sugar water, but making sure no one else could profit from the "secret formula."

May 1886. Atlanta.
John Pemberton. A veteran who had been gravely wounded in the Civil War. After his injury, he became addicted to medical morphine. To break the morphine addiction, he experimented daily in his basement with various plant extracts, searching for a nerve tonic that "could keep a person clear-headed without relying on alcohol or morphine." His tools: a copper kettle, a wooden spoon, and a pharmacy scale.
He mixed several ingredients: coca leaf extract (containing trace cocaine), kola nut extract (containing caffeine), plus caramel, phosphoric acid, and lemon essential oils. Deep brown. Sweet with a bitter edge.
He carried the copper kettle down to Jacob's Pharmacy, the largest in Atlanta — and said: "Sell it for me. Five cents a glass."
Total sales for the first year: 25 gallons. Converted to today's measures — roughly a thousand glasses. Almost no one bought it.
Pemberton's accountant, Frank Robinson, wrote the trademark in Spencerian script — Coca-Cola. Then Pemberton's health began to deteriorate. He was so poor that he sold off bits and pieces of the formula to different people — and by the time he died in 1891, he was nearly penniless.
The person who truly made Coca-Cola live was not its inventor. It was Asa Candler.

After Candler bought the formula, the first thing he did was not expand production — he stopped selling it as "medicine."
He stripped off every label that said "cures headaches" or "relieves fatigue" and changed the "brain tonic" into — "a thirst-quenching beverage. Refreshing. Delicious."
Then he distributed free redemption coupons at Atlanta pharmacies — "Take this slip of paper to any soda fountain and get a free ice-cold Coca-Cola."
Next, he shifted the marketing focus from efficacy to sensory conditioning — a blend of orange blossom, vanilla, and cinnamon aromas. The soda fountain had to have a white marble countertop and a server in white gloves. He wanted people, at the mention of Coca-Cola, to think not of "medicine taste" but of "cold air," "ritual," "the scent of summer."
Another pivotal decision by Candler came in 1899, when he signed over the bottling rights for Coca-Cola to two young lawyers from Chattanooga — practically for free. The price was a symbolic $1 — while the syrup would be supplied exclusively by him, and all the profit would stay inside the syrup company. Those two lawyers and their franchisees would be responsible for all the bottling equipment, glass bottles, trucks, and refrigerated coolers.
This created a peculiar business structure: the Coca-Cola Company only makes concentrate syrup — the highest-margin, lowest-capital-investment part of the business. Every single bottle of sparkling Coca-Cola on earth is diluted from that concentrate. The bottlers shoulder most of the fixed assets and labor.
This was Candler's ultimate design — forever hold the formula, forever sell only the syrup. Offload the heaviest asset burdens onto the franchisees — and then use marketing and brand power to make consumers transmit the pressure back to the bottleneck: if a bottler refused to stock Coca-Cola, consumers would get angry.

The marketing of the "secret formula" began after 1919 and reached mythological status under Robert Woodruff. He locked the formula in a vault at SunTrust Bank in Atlanta — one of the most effective marketing operations in history. He publicly declared that "only two people in the world know the full formula." Every time a consumer cracked open an ice-cold Coke, they felt like they were drinking a liquid encoded with a cipher.
Today, there is no one who genuinely couldn't replicate the beverage through chemical analysis. But the "secret" itself has become the brand. It is the only beverage on earth that has turned its ingredient list into an urban legend.

Now stop for a moment.
Inside your product, is there something "others could analyze but you will never publicly say out loud" — a mysterious element — that makes your customers feel they're buying not a product, but a cipher?
If not — have you been explaining everything a little too clearly?

Three things.
First: you're not selling a product — you're selling a conditioned reflex.
Coca-Cola's advertising never talks about "how much sugar, how much carbonation." It does only one thing: bind the color red, Santa Claus (Coca-Cola redefined Santa's red-robed image in 1931), polar bears, and the ice-cold glass bottle together with "happiness." Every element is a neural chain embedded in memory.
In your customer's brain — which non-functional emotion is your brand bound to so tightly that the two are indivisible?
Second: only make the syrup.
Coca-Cola doesn't do bottling. Doesn't do transportation. Doesn't do retail. It only does the lightest, highest-margin link — the one where the technical content lives inside "taste memory" — the syrup.
In your business model, which links are you hauling heavy boxes for someone else — instead of making your own "concentrated syrup"?
Third: turn a trade secret into a public myth.
Coca-Cola's "formula secrecy" is fundamentally a branding act, not a security measure. Cracking the chemical formula costs only a few hundred dollars. But inside that vault — it's priceless.
Do you have a "craft story" — not deliberate mystification, but a "sense of mystery" your customers can tell their friends about?

Asa Candler died in 1929. By then, Coca-Cola had spread from the basement of an Atlanta pharmacy to 78 countries around the world. In his final will, he donated the bulk of his fortune to Emory University and the Methodist Church.
When he bought the formula, people thought he'd paid $2,300 for a copper kettle and a pot of brown sugar-water — far too expensive.
Then he added one ingredient to that pot of water that no one else had — a marketing system wrapped in mystery that enveloped the entire globe.

← Previous Contents Next →
Chapter 31 Procter-&-Gamble

1837, Cincinnati. An English immigrant candle maker and an Irish immigrant soap maker — their delivery wagons shared the same cart path. They married a pair of sisters — and then their father-in-law called them to the table and said: "You are family now. Merge." And so Procter & Gamble was born. Today, P&G has a market cap of $337 billion, and more than 5 billion people on earth use a P&G product every single day.

Not brand marketing. Not category monopoly. Not advertising budgets. Not channel control.
Just William Procter and James Gamble. One made candles, the other made soap — both products, in 1837, required animal fat as raw material. The first year after merging, they simply shared a warehouse. Then they realized — if they didn't merge, they'd end up competing for raw materials later. If they merged, they could squeeze suppliers together.

Cincinnati in the 1830s was "Porkopolis" — the largest pork processing center in America. After hogs were slaughtered, the meat was cured into ham and bacon and shipped to the East Coast, while the lard and tallow — originally waste products — were collected by local soap and candle factories. Procter and Gamble's business model was brutally simple: buy tallow → boil soap / pour candles → sell to pioneering families and shops heading west. But the quality of the tallow was wildly inconsistent — sometimes the lard went rancid, and you came out of the bath smelling worse than when you went in.

James Gamble was a chemistry fanatic. He built a laboratory behind the soap workshop — in reality, just a wood-planked cubicle with a scale and a thermometer — and tested different ratios of tallow, lye, and rosin every day. His breakthrough happened entirely by accident: one evening in the 1860s, the mixer operator forgot to shut down the machine and let a batch of soap slurry churn overnight. The next morning, Gamble came in and saw — the fats had separated more thoroughly than ever before. That batch of soap floated in cold water — and its lather was incredibly fine. This was Ivory Soap — "It floats — 99.44% pure." That accident later became a deliberate procedure, with P&G engineers "intentionally" introducing an extended mixing cycle into the process.

Then P&G invented modern brand management. In 1931, a young advertising manager named Neil McElroy wrote an internal memo — proposing that the company treat each brand as an independent business, with a "brand manager" vested with full profit-and-loss authority over that brand. That memo later became the organizational blueprint for every consumer goods company on earth.

Through the 1930s to the 1960s, P&G mass-invented products that consumers couldn't live without: Tide (the first synthetic laundry detergent — no more scrubbing soap bars on a washboard), Crest (the first fluoride toothpaste), Pampers (the first disposable diaper), Head & Shoulders (the first anti-dandruff shampoo). In every category, P&G was the company that was "not the first — but the first to get it right."

Three things.
First: build consumer goods with chemical logic — not marketing logic.
Gamble improved soap formulas using burettes that measured the acid value of tallow. The core of Crest toothpaste was the clinical data on stannous fluoride from the American Dental Association — P&G bought the patent rights. Tide's formula was built on "synthetic surfactants" — not natural soap bases.
In your industry — is there still someone competing on "advertising slogans" — while you have an opportunity to strike from a lower dimension with "molecular formulas"?
Second: the brand manager system — slice a big company into small units.
One brand, one manager, one P&L. P&G doesn't rely on a CEO to manage 300 brands — it relies on 300 "mini-CEOs" competing with each other for headquarters' attention and capital.
In your company — is the boss managing every product, or does every product have "its own boss"?
Third: industrial accidents shouldn't change the process — they should change the product.
Soap churned overnight was not SOP. A floor-cleaning compound that didn't taste bitter was not SOP. But P&G's engineers were trained to hunt for new product opportunities inside "anomalies" — rather than filing them away as incident reports.
Have any of your "operational mistakes" — ever become a new product?

William Procter's grandson, William Cooper Procter, introduced an extraordinarily rare policy in 1923 — "Every employee receives an annual profit-sharing bonus, with the bonus amount tied to company profit." He was among the earliest American industrialists to champion "profit sharing." He said: "If you give a man only a fixed wage — by the 31st day, he becomes your cost."

← Previous Contents Next →
Chapter 32 Applied-Materials

1967, Santa Clara, California. A chemical engineer named Michael McNeilly — coming out of the semiconductor materials division at Texas Instruments — built a small lab in his own garage and began using chemical reactions to "grow" an exquisitely uniform insulating film onto the surface of silicon wafers. He named the company Applied Materials. The name meant: he doesn't make chips — he makes "the equipment used to manufacture chips." Today, Applied Materials has a market cap of $332 billion and is the largest supplier of semiconductor manufacturing equipment on earth — nearly every chip in your hand has passed through one of its machines.

Not chip design. Not wafer fabrication. Not Moore's Law. Not TSMC.
Just the physics of "deposition, etching, and polishing" — stacking atoms layer by layer onto a silicon wafer, then using plasma to blast away the unwanted portions — until that silicon surface reveals billions of transistors, each only a few dozen atoms wide. What Applied Materials' machines do is "the addition and subtraction of atomic layers."

Michael McNeilly was born in California in the 1930s and earned his chemical engineering degree from Stanford. At TI, he was responsible for semiconductor manufacturing processes — in that era, chip fabs built their own equipment, and every machine was a custom one-off. McNeilly saw that this model of "every chip fab builds its own equipment" could never scale — as transistor density began doubling in lockstep with Moore's Law, the complexity of the processes would outstrip the equipment-development capacity of any single semiconductor company's internal R&D. He believed there had to be an independent company — one that only made equipment — and sold it to every chip fab.

In 1967, he founded Applied Materials. Its first product was an epitaxial reactor — at temperatures of over one thousand degrees Celsius, silane gas was flowed across the surface of a silicon wafer, causing gaseous silicon atoms to "grow" onto the wafer surface layer by layer, forming a perfect crystal lattice. The machine was sold to several of Silicon Valley's early chip companies.

But what truly turned Applied Materials into a behemoth was — during the 1980s and 1990s, it expanded its product line from a single epitaxy machine into physical vapor deposition (PVD — using high-energy particles to "bombard" metal atoms off a target material and deposit them onto the wafer surface), chemical vapor deposition (CVD — using gases that react on the wafer surface and leave behind a solid thin film), chemical mechanical polishing (CMP — using nano-scale slurry to flatten copper interconnect layers), and plasma etching. It essentially came to own half the realm of chip manufacturing — all the critical equipment for "adding material, subtracting material, and flattening" on the wafer.

During the AI chip explosion from 2023 to 2025 — Applied Materials became one of the single most indispensable equipment suppliers to TSMC, Intel, and Samsung. Producing a single NVIDIA H100 GPU chip requires over a thousand process steps — roughly half of those steps use Applied Materials equipment.

Three things.
First: see clearly, as early as 1967 — that chip fabs building their own equipment could never keep pace with Moore's Law.
McNeilly wasn't a semiconductor physicist — he was a chemical engineer. He calculated from the rate of chemical reactions that if process complexity doubled every 18 months, no single company's internal R&D could independently cover the full suite of new equipment required.
In your industry — are you "internally building" a component that will sooner or later exceed your own R&D capacity — and should you spin it off into an independent industry?
Second: Applied Materials doesn't sell equipment — it sells "yield at the atomic scale."
A chip fab spends thousands of dollars in cost per wafer — and if Applied Materials' precision control can lift yield from 80% to 90%, each wafer produces hundreds more saleable chips. That value is dozens of times the selling price of the machine.
Your customer — is buying your product to save costs, or to improve output yield? The pricing ceiling for the latter is far higher than for the former.
Third: the equipment behemoth of the AI era — doesn't make chips but earns more than the chip fabs.
Applied Materials bears no chip-inventory risk, no design-failure risk, no risk of failed R&D on advanced process nodes. But to every chip fab that has placed those risky bets — it sells the indispensable equipment for "manufacturing atoms."

Michael McNeilly retired from the board in 2005. Outside Applied Materials' headquarters in Santa Clara — in front of the building stands a stainless-steel reaction chamber, removed from the very first epitaxial reactor. A layer of faint purple polysilicon deposition film, never fully cleaned off from decades ago, still clings to the inner wall — shimmering with iridescence in the sunlight.

← Previous Contents Next →
Chapter 33 Palantir

After 9/11, a Stanford philosophy PhD and a member of the PayPal Mafia sat down together and made a decision: we are not building an AI assistant. We are building a counterterrorism platform to "find the terrorists." And then the U.S. government became their first — and for well over a decade, their only — real customer.

Not consumer tech. Not enterprise SaaS. Not a platform ecosystem.
Just Peter Thiel and Alex Karp. 2003 — sixteen months after 9/11. Thiel handed Karp a check and said: America's intelligence agencies need software that can take all the fragmented information — phone records, bank transactions, surveillance footage, informant reports — and connect it inside a relational reasoning engine. This software isn't meant to save you time. It's meant to find the next person who wants to fly a plane into a building.
And then they spent 15 years bound to the government, with virtually no commercial customers. Every venture capitalist shook their head and said: "This company will never be commercializable."
Today, Palantir has a market cap of $322 billion. It is one of the most controversial, and yet also most singular, technology companies on earth.
Let me tell its story first. When I'm done, we'll talk about why "following one impossibly difficult customer for a lifetime" may be more commercially valuable than "chasing every customer."

Alex Karp. Born in 1967, grew up in California. Undergraduate degree from Stanford; his PhD was in social theory from Goethe University Frankfurt in Germany — his dissertation was on "how the individual resists authority." He does not have an easy personality — he will occasionally argue with the air in German. He holds a black belt in karate. His early friends describe him as "the most human-porcupine-like mammal you will ever meet."
Peter Thiel was a co-founder of PayPal and Facebook's first outside investor. In 2002, PayPal sold, and Thiel began turning over a question in his mind: Silicon Valley could make it possible for a middle-schooler to video-chat with strangers across the world — but 19 terrorists had planned and executed 9/11 in total invisibility, scattered across the fragmented file cabinets of the FBI, the CIA, and the Department of Homeland Security. Could that systemic gap be closed with software?
He recruited Karp. Thiel was the largest shareholder; Karp was the CEO. The company's name wasn't drawn from business vocabulary — it came from Tolkien's world of Middle-earth. A Palantír is a "seeing stone" of that world — through it one could perceive all connections, but its user had to possess great wisdom, or it would consume them.

For the first five years, Palantir had only government clients. CIA. FBI. Department of Defense. Immigration and Customs Enforcement. What these agencies had in common: data scattered across hundreds of mutually incompatible databases. Phone records in one system, suspicious bank transactions in another, border crossing records in a third — and the only way to connect any of them was a human analyst hitting Ctrl+F in Excel.
Palantir built the Gotham platform — pouring all of that data into a single semantic graph. It wasn't "AI prediction" — it was "human analyst + AI-assisted correlation" — preserving the human's decision-making authority at the most critical judgment points, but compressing the speed of information retrieval from days down to seconds.
This drew enormous social controversy. Privacy organizations sued Palantir — claiming it helped the government analyze citizen data without any warrant. Karp never dodged this topic. His response was an almost philosophical defense: "We have to choose. We either use software to protect free societies — or we let organizations that have no democratically elected government use software more advanced than anyone else's to destroy those societies."
This kept commercial customers away from Palantir for many years. A shadowy software company bound to the CIA — for corporate enterprises, that was no cuddly IT vendor.

From 2012 to 2014, Karp decided Palantir needed to enter the enterprise market. He named an analytics platform Foundry — the underlying logic was the same: a giant corporation — Airbus, for example — also had thousands of discrete Excel spreadsheets and SAP instances internally. Foundry turned them into a queryable human+AI data ontology.
But sales were brutally difficult. Many commercial customers felt that "spy-grade tools" weren't suited for consumer-facing companies. Their sales cycles in many cases exceeded 12 to 18 months. And Karp insisted on never discounting.
"If we lower the price — that's the same as saying we don't believe in our own product."
2020, Palantir went public. During the investor Q&A session on the IPO roadshow, Karp did something that violated every investment-banking rule. He didn't read from a pre-written script. He looked at the screen and said something I remember roughly as: "We are not optimizing for short-term shareholders. If you're planning to sell next month — please go buy the SaaS company next door. What we're doing here is a 20-year project."
The first year of trading, the stock was flat.
2023, the generative AI explosion. Major corporations and defense agencies suddenly discovered — AI requires a clean, fully connected, ontologically consistent data layer in order to work. And this was precisely the thing Palantir had been building for nearly 20 years.
2024-2025, Palantir's AIP (Artificial Intelligence Platform) was adopted by the U.S. military, healthcare systems, and large manufacturing groups in rapid succession. Its order book detonated. Wall Street finally began to pull it apart and study it — and discovered: this company that "could never be commercialized" was already running command-analysis systems for the defense ministries of 20 NATO countries, while simultaneously generating steady commercial revenue growth — and its average customer partnership duration was 14 years.

Now stop for a moment.
Do you have a customer you've served for more than five years — one that's been a loss leader the entire time, one you've been grinding against — but whose demands have honed your product into a machine that can operate in "the harshest possible environment"?
Karp never did this for a quick buck. He spent ten years serving only the U.S. government — the most demanding, least price-sensitive, impossible-to-please customer in the world. And then he discovered — when your system can already save lives on a battlefield, selling it to banks and aircraft manufacturers is no longer a problem.
Are you willing — to wrestle with the hardest customer for ten years, in exchange for a standard no one else can ever catch up to?

Three things.
First: the product didn't come first — the extreme environment did.
Before Palantir's software had a single feature, it was already placed in the post-9/11 counterterrorism environment of "if this doesn't run, people die." Most startups' products are born in air-conditioned conference rooms and on whiteboards. Palantir was born in battlefield map rooms and mission briefings before nighttime raids.
What environment was your product born into? "Gentle" — or "lethal"?
Second: the data ontology > AI.
Karp has been saying one thing over and over: "Artificial intelligence is only a thin surface layer. The truly valuable thing is the ontology layer beneath it — the layer that can accurately map all data onto real-world objects." While other companies were training large models, Palantir was building data connections — creating a semantic ontology for every organization.
Is your company building AI's "spire" — while ignoring the foundation that requires a decade to build beneath it?
Third: do not apologize for your political stance.
Palantir lost massive commercial clients for years because of its government contracts — Karp never once apologized for it or pivoted away. He said: "The U.S. government and its allies need the best software — and others are unwilling to provide it. We provide it." That stance cost him enormously in commercial terms. And then it attracted every top-tier engineering talent who shared that stance — engineers who wanted their code to stop bullets.
Do you have a firm stance — one that "makes some people leave but makes others fight to the death to work for you"?

In a 2023 interview, Alex Karp was asked, "What is Palantir actually doing?" His answer made even the reporter pause for three full seconds before he continued speaking:
"We are building the data layer for the operating system of Western democratic society."
He is not a normal CEO. Palantir is not a normal company. And that $322 billion worth of not-normal — is its unique moat.

← Previous Contents Next →
Chapter 34 Home-Depot

In 1978, two 49-year-old men were fired by the company they had built. They sat in a coffee shop in Atlanta. One of them said: What if we opened a hardware store — better than the one that just kicked us out?

Not a youthful startup. Not a Silicon Valley myth. Not venture capital. Not an IPO cash grab.
Just Bernie Marcus and Arthur Blank. Two middle-aged men. Fired on the same day from the Handy Dan home-improvement chain — because they lost a power struggle with the parent company's majority shareholder. That year, Bernie was 49. Arthur was 49. Their prime years. No pension. Both were ordinary people who had just lost their jobs.
And then they turned their rage into the largest home-improvement retailer in human history — Home Depot, market cap $310 billion today.
Let me tell their story first. When I'm done, we'll talk about why "getting fired" is sometimes the best business plan you will ever encounter.

April 1978. San Diego, California.
The CEO of Handy Dan's parent company, Daylin, was a man named Sandy Sigoloff. He ran a medical chain. He didn't understand hardware retail. But he controlled the board. He fired Bernie Marcus — the company's most successful regional general manager — for "philosophical differences." And Arthur Blank — the CFO — was marched out the same day.
That night the two men met at a coffee shop. Bernie had a black coffee in front of him. Untouched.
"What do we do now?"
Arthur: "We pour everything we've ever learned into it — and build a company where no employee ever has to go through what we went through today."
They pooled every dollar of their savings, then found a New York venture capitalist named Ken Langone. Langone listened for forty minutes and said: "What you're going to open is a massive space unlike anything that's existed — with more than 25,000 different hardware items — and every salesperson on the floor must be a genuine carpenter or electrician, not a cashier."
He put in $2 million in seed funding.
June 1979, the first Home Depot opened in Atlanta. It occupied 55,000 square feet — the average traditional hardware store at the time was only 5,000 square feet. Walk inside, and supermarket-style shelves were stacked twenty feet high. Every pipe, plank, wire, and paint you could imagine — including tools you didn't know you needed but genuinely did.
The most critical element was the "master-tradesman sales force." Every store had to maintain a certain ratio of licensed electricians, plumbers, and carpenters on the floor — because the customers coming to buy tools didn't necessarily know how to use them. They needed someone to teach them. This was a free home-improvement class, happening in every aisle. And the people who got the lesson ended up buying more materials and tools.
1981, Home Depot went public. Those "big boxes" spread like wildfire from Atlanta to Florida, Texas, California — often, after the fourth store opened in a city, the local independent hardware shops simply couldn't survive.
This infuriated a lot of people. The main street of every small town had a hardware store that had been in the family for three generations. When Home Depot moved in — those tiny shops closed within months. Community protests. Newspaper headlines. Local politicians filing lawsuits — Bernie and Arthur never stopped.
Because they had lived through one thing: the day they were fired, not a single competitor called to ask, "Are you okay?" So the entire industry owed them no emotional debt.

In 1997, Bernie and Arthur officially retired — and handed the CEO role to Robert Nardelli, recruited from GE. Nardelli did something that every business school has since written up as a case study — he gutted Bernie and Arthur's "master-tradesman culture." To cut costs, he replaced licensed electricians with part-time high school kids. Employee morale fell through the basement floor. But short-term profit margins rose.
Bernie later said on many occasions that this was "one of the biggest regrets of his career."
But here's the interesting part — Home Depot's grassroots culture was so strong that it survived this wave of management cultural destruction. Nardelli left in 2007. Frank Blake took over and gradually restored the tradesman hiring policy. Today, the orange aprons are still in every aisle — someone there can still help you figure out how many bags of cement you need.

Now stop for a moment.
Have you ever been fired? Have you ever harnessed the fury of being rejected — and turned it into fuel for declaring war on an entire old industry?
Bernie wasn't opening a supermarket. He was using 55,000 square feet of product lines to exact revenge, on behalf of himself and his best friend, for that afternoon of humiliation — against the entire old model. The most enduring commercial drive often comes not from foresight — but from humiliation.

Three things.
First: biggest = cheapest.
Home Depot's scale could command supplier prices so low — squeezed down to a level no local hardware store could match. And then it used the money saved to hire actual licensed tradesmen — erasing the small shops' last shred of advantage: "We know more than the big box."
What do you use to crush the little guys? Not prices — but "a fixed investment others can't afford but you can amortize with scale" (like full-time master tradesmen).
Second: teach the customer how to use it — and they'll buy more.
This is a counterintuitive but endlessly verified iron law of retail. Most people aren't unwilling to do home improvement — they just don't know how to start. A tradesman teaches them for ten minutes — and for the next half hour, their shopping cart fills up.
Do you have a "teach the customer how to use it" touchpoint — that could directly triple the average transaction value?
Third: bake the fury of being fired into the company constitution.
The first value Bernie and Arthur established at Home Depot was — "Respect for all people." Not because they'd read a leadership book. Because they themselves had been rudely thrown out, and they knew exactly what that felt like.
Is your company's number-one value — drawn from the single biggest event of your life?

Bernie Marcus is 95 years old. He has since donated virtually all of his money — tens of billions of dollars. When asked, "What decision do you consider your most successful?" — he said:
"I thank the CEO who fired me in 1978. Without him — I would never be who I am."

← Previous Contents Next →
Chapter 35 General-Electric

Thomas Edison was the founder of GE. But his company later did something he could never have imagined — a hundred years after his death, GE split into three companies. And the filament inside Edison's original incandescent bulb, in a sense, is still burning.

This is not about Edison inventing the light bulb. The story is more complicated than the light bulb.
In 1892, Edison merged his Edison General Electric with its competitor Thomson-Houston — forming General Electric. This was the fruit of the first large-scale industrial consolidation in late-19th-century America. Edison was a technical genius, but he was no organizational genius. His workshops didn't even set up cost centers.
It was Charles Coffin of Thomson-Houston — who was the true founder that turned GE into a modern corporation. The management systems and budgeting frameworks he established were later emulated by every industrial giant that followed.
And then, over the next 130 years, GE experienced more cycles of "big-bang" expansion and "major-surgery" contraction than any other company. Its history is a concentrated record of every choice American industry made from the Gilded Age to the digital age.

Charles Coffin ran GE from 1892 to 1912. He didn't make specific inventions. But he did one thing — he turned GE into a "company of professional managers" — where all decisions were based on numbers, not an inventor's intuition.
And then Jack Welch arrived.
In 1981, Welch took over as CEO of GE. Over the next 20 years, he ran an experiment on a global scale: in every industry, if any GE business could not be number one or number two in its market — sell it or shut it down. He transformed GE from a sprawling manufacturing conglomerate with hundreds of production lines into a highly focused industrial-financial hybrid.
During the Welch era, GE was worshipped by MBAs as the god of management. GE's market cap went from $14 billion in 1981 to $410 billion when he retired in 2001 — the highest in the world.
But he also planted a time bomb: GE Capital — the financial engine. Welch discovered that lending against industrial assets earned more, and faster, than industrial manufacturing itself — and so more than half of GE's profits came from finance. This was a game of using an AAA-rated industrial credit rating to issue loans.
2008, the financial crisis. By then, GE Capital was sitting on hundreds of billions of dollars in loans, leases, and real-estate exposure. After Lehman Brothers collapsed, GE's short-term financing evaporated overnight. This company, once considered invincible, was forced to go begging to Warren Buffett — Berkshire Hathaway injected $3 billion (and received highly preferential stock rights in return).
From that day forward, GE entered a de-leveraging and slimming-down process that lasted over a decade: selling off GE Capital, selling appliances, selling biopharma, selling oil and gas equipment — until 2021, when it announced the final plan: break what remained of GE completely into three independent companies — GE Aerospace (aircraft engines), GE Vernova (energy), and GE HealthCare (medical).
The all-encompassing "management platform" that Coffin had built — was dismantled. Edison's incandescent bulb no longer burns. But GE aircraft engines still blaze blue flames beneath the wings of every Boeing and Airbus.

Now stop for a moment.
Over GE's 130 years — expansion, diversification, financialization, crisis, dismantling — every step "looked right at the time." When Welch was treated as a god, no one questioned him. And then the bomb he planted detonated seven years after he retired.
Your current growth engine — does it have a timer you yourself cannot see?

Three things.
First: the myth of industrial diversification can blind everyone.
GE once "proved" that a single management team could simultaneously run medical devices, aircraft engines, television stations, insurance, and home mortgages — this was shown to be one of the greatest illusions in the history of management. Later, the market stopped paying a premium for it — investors learned to discount the value of diversified conglomerates.
In your business portfolio — how many pieces are there that "you thought would synergize but actually have nothing to do with each other"?
Second: finance can very easily turn an industrialist into a gambler.
The vast scale of GE Capital gave GE years of excess profits — and then required even more years to pay back every single risk on that profit statement. Finance is sweet poison — the pleasure at the moment of consumption is intense, but the digestion is lethal.
How close is your company to "replacing product competitiveness with financial leverage"?
Third: breaking up is not failure — it is a form of clarity.
GE's final dismemberment was not because the company had no value — it was because the market assigns higher value to pure-play businesses. GE Aerospace, standing alone, has a market cap higher than the original GE had before the breakup. Admitting that a complex structure has failed — and voluntarily dismantling it — requires far greater courage than building it in the first place.

From Edison's light bulb to the pinnacle of human aviation engines — GE survived every technological explosion. It was not killed by competitors. It was nearly devoured by finance. And then it redeemed itself through a form of self-amputation.

← Previous Contents Next →
Chapter 36 General-Electric

1892, Schenectady, New York. Two inventors held a tangled knot of patent lawsuits in their hands — you sue me for infringing on the electric light patent, I sue you for stealing AC power transmission. Behind one stood J.P. Morgan; behind the other, the Vanderbilt family. Morgan summoned both sides into his study and said one sentence: "From this moment on, neither of you exists — there is only one company, called General Electric." And so Thomas Edison's Edison Electric — and Elihu Thomson's Thomson-Houston Electric — merged. Today, GE has a market cap of $307 billion.

Not the Industrial Revolution. Not military contracts. Not financial engineering. Not GE Capital.
Just that merger in the study. But what truly allowed GE to survive for 130 years was not that merger — it was that GE built a "self-reinvention machine": every time the industrial era shifted, GE used the profits from the previous era to buy a ticket to the next.

After Edison lit the first practical incandescent bulb in 1879, he founded Edison Electric — but he was a diehard devotee of direct current. His competitors George Westinghouse and Nikola Tesla had already proved that alternating current could transmit efficiently over long distances — whereas DC required a power station every few kilometers. The merger that Morgan orchestrated in 1892 had a second purpose — to squeeze Edison himself entirely out of management, because he was obstructing the adoption of AC. Edison left GE and never returned.

And then GE found Charles Steinmetz — a hunchbacked German immigrant mathematician who stood under five feet tall, known as "the Wizard of Schenectady." His breakthroughs in AC theory made GE the world's largest supplier of AC power generation and transmission equipment. From there, GE swallowed the entire electrical product line, from light bulbs to ovens to refrigerators to gas turbines to jet engines — from generation to consumption, it devoured the entire value chain of electrification, top to bottom.

But what truly made GE the "living fossil of management history" was Jack Welch — who became CEO in 1981, at a time when GE was one of America's largest companies, but slow and lumbering. Welch reordered every business unit inside GE according to one standard: "Be number one or number two in your industry — or be sold." He sold off more than 200 business units, laid off 100,000 employees, and was dubbed "Neutron Jack" by the media. Then he inflated GE Capital — using the manufacturing arm's AAA credit rating to borrow in the financial markets and invest in financial services. At its peak, GE Capital contributed 50% of the group's total profits. The 2008 financial crisis smashed GE Capital's profit myth back to reality — the government injected $140 billion in emergency loans to keep it afloat.

Subsequently, through the 2010s, GE entered a painful period of slimming down — selling off appliances, selling NBC Universal, selling off most of GE Capital, and finally, in 2024, splitting GE itself into three independent public companies: GE Aerospace, GE HealthCare, and GE Vernova (energy). That thing Morgan had pieced together in 1892 — was taken apart 132 years later.

Three things.
First: don't love an invention — love electricity.
Edison loved direct current. Morgan loved "electricity that could be sold." The former lived inside the technology. The latter lived above it.
What you're holding onto right now — is it "direct current" or "electricity itself"?
Second: either be number one or number two — or don't play.
Welch's brutal ranking produced enormous short-term profits — but it also produced a pathological obsession with "rankings," leading GE Capital to use balance-sheet arbitrage to sustain the appearance of growth. True number-one status — is chosen by customers, not manufactured on a financial statement.
Is your number-one ranking — a vote from your customers, or a number calculated on your spreadsheet?
Third: the empire may be dismantled — but the engines remain.
GE disintegrated — but GE Aerospace engines fly on 70% of all global flights. GE HealthCare's CT scanners and MRI machines run in hospitals around the world. GE Vernova's wind turbines and gas turbines generate power. What survived was not the brand empire — it was the engineering legacy.
If your company were dismantled today — what would survive as its "legacy"?

Charles Steinmetz died in 1923. The contract he signed with GE during his lifetime was absurdly lopsided — no fixed salary, only a massive annual retainer, with the freedom to take vacation at any time and the right to refuse any assignment. On his tombstone was not engraved "Chief Engineer of GE" — but "Contributor to the Theory of Alternating Current."

← Previous Contents Next →
Chapter 37 Morgan-Stanley

September 16, 1935, 2 Wall Street — amidst the partners of J.P. Morgan, someone suddenly threw his napkin down onto the table. The gesture came from Henry Sturgis Morgan — grandson of J.P. Morgan. He stood up and announced to the partners of the Morgan dynasty: he and three other partners were leaving the House of Morgan — to open a new investment bank. This was because the Glass-Steagall Act of 1933 mandated that commercial banks could no longer engage in investment banking. Either split up, or break the law. And then Henry Morgan chose a third path — he didn't wait to be broken apart. He walked straight out the door, went around the corner from 2 Wall Street — and opened Morgan Stanley. Today, Morgan Stanley has a market cap of $296 billion.

Not financial-instrument innovation. Not government connections. Not trading genius. Not capital scale.
Just Henry Morgan. A man who watched the financial empire his grandfather had built be forced to choose between "commercial banking" and "investment banking." He didn't weep. He turned around, walked one block — and from that day forward, Morgan Stanley devoured a huge share of the investment banking business that had originally belonged to J.P. Morgan, across securities underwriting, M&A advisory, and high-net-worth wealth management.

Henry Morgan was born in 1900, into a family where he didn't need to "start a business." His grandfather, J.P. Morgan — during the Panic of 1907, single-handedly locked the doors on the debtors of several of Wall Street's largest banks, demanding that all bankers sign a joint rescue agreement, and was regarded as America's "unofficial central banker" in the era before the Federal Reserve. Henry grew up in that environment — but he chose a strange path: he didn't stay on the family's commercial banking side. He went to the investment banking half.

When he left J.P. Morgan in 1935, he took with him partners Halsey Stewart, Robert Stanley, and several others — Stanley's name later became the "Stanley" in "Morgan Stanley." On its first day in business, the firm had twelve clients — six were legacy clients from the J.P. Morgan orbit, and six were newly recruited. Henry didn't compete on price — he competed on: "If you trusted my grandfather's balance sheet — now you can continue to trust my people."

For the decades that followed, Morgan Stanley was intensely elitist — its employees were exclusively Ivy League white men; it did not raise outside equity, did not disclose its financials publicly, and did not open branch offices. It was called "the white-shoe firm of Wall Street" — the implication in its marketing was: "We don't do business with everyone — we only do the most complex business with the wealthiest clients." In 1986, Morgan Stanley finally went public — and then began its globalization.

During the 2008 financial crisis, Morgan Stanley — like Goldman Sachs — was not a commercial bank, but under a special dispensation from the Fed and the Treasury, converted into a bank holding company to access emergency government loans. During the most panicked days, it was mere hours away from bankruptcy. And then it saved itself — by selling convertible preferred stock and accepting a capital injection from Japan's Mitsubishi UFJ.

Three things.
First: where regulation cuts you apart — that's the starting point of a new company.
The Glass-Steagall Act forced Morgan Stanley into existence. The law can tear a company apart — but that company's culture, its clients' trust, and its "way of doing things" will continue as before, under a new sign.
The fissures carved into your industry by regulation — are they the starting point of your next company?
Second: not doing everything for everyone — that itself is a strategy.
Morgan Stanley maintained its "elitist exclusion" strategy for 50 years. It caused them to miss out on mass-scale businesses like retail brokerage and credit cards — but in the highest-margin zones of M&A advisory and IPO underwriting, it secured an unassailable first-tier position that no one could steal.
What large market have you refused — in exchange for what unshakable position?
Third: hours from bankruptcy — and then Japan saved it.
Morgan Stanley didn't survive 2008 through cleverness — it survived because of a $9 billion check written by Mitsubishi UFJ at the very peak of the panic. And the reason the Japanese were willing to write that check — was that, twenty years earlier, when Japan's bubble economy burst, Morgan Stanley had not pulled out of the Japanese market.
In the long-term trust you've built with your clients and partners — is there a "life-saving check, 20 years from now," already on deposit?

Henry Morgan died in 1982. Throughout his life, he insisted on one rule: "A partner of Morgan Stanley must personally attend every single client board meeting." He said: "If you are not in the room where the client's decisions are made — you are merely selling a piece of paper."

← Previous Contents Next →
Chapter 38 Philip-Morris

1847, Bond Street, London. A cigarette-shop proprietor named Philip Morris opened a small retail store. He wasn't selling "cigarettes" — cigarettes hadn't even been invented yet. He sold hand-rolling tobacco and imported Turkish pipe tobacco. Outside his shop hung a handwritten wooden sign: "Philip Morris — Tobacconist & Barber." Because he ran a barbering business at the same time. Today, Philip Morris International has a market cap of $291 billion and is one of the largest tobacco companies on earth. Marlboro, under its umbrella, is one of the best-selling consumer brands in human history.

Not addiction. Not government acquiescence. Not black markets. Not consumer ignorance.
Just one very strange fact: when Marlboro was born in 1924, it was a women's cigarette — the ad copy read "Mild as May." Before the 1950s, Marlboro was the weakest product in the Philip Morris lineup — its market share was under 1%. Then an advertising man named Leo Burnett — in a conference room in Chicago — showed the president of Philip Morris a photograph of a cowboy. He said: "Take that cigarette off the woman and give it to this man." And then Marlboro became "the symbol of masculinity" — sales shot up 3,000% in a single year.

But what I want to talk about is something more fundamental: at every stage, Philip Morris did the same thing — "when the old market is about to die, use brand power to plant your flag early in the new market." In the 1990s, when Western cigarette markets began to decline, it accelerated its expansion into Eastern Europe, Asia, and Africa. It acquired local tobacco companies in the Czech Republic, Lithuania, Kazakhstan, Indonesia, and the Philippines — not by dumping Marlboro directly into those markets, but by preserving local brands as distribution pipelines, then seeping Marlboro along those pipelines. To this day, the vast majority of Philip Morris's revenue comes from markets outside the United States.

And then in 2014 — Philip Morris launched IQOS (a heat-not-burn tobacco device) and rolled it out aggressively in the Japanese market. It declared that its vision was to "ultimately end traditional cigarettes." The profound paradox behind this is — this company's existence depends on selling cigarettes, and yet its stated goal is to eliminate cigarettes.

Three things.
First: the weakest marginal product — swap its packaging and its symbol — can become the strongest.
Marlboro's journey from women's cigarette to Western cowboy is the most extreme reversal case in the entire history of 20th-century branding. The core operation was not changing the formula — it was changing the image the consumer saw in the mirror of his own mind when he looked at the pack.
Is your product — trapped in a "first impression" it cannot escape?
Second: before the regulations ban you — shift your growth to markets the regulations haven't reached yet.
While Philip Morris was besieged across the West by anti-smoking campaigns, advertising bans, and Marlboro litigation, it used dozens of local brands to build un-extractable distribution bases in emerging markets.
Was your most important growth — already planted elsewhere before your industry got "banned"?
Third: invent the substitute that kills you — it's better than letting someone else kill you.
The logic behind Philip Morris pushing IQOS and e-cigarettes was not a moral transformation — it was "if we don't end ourselves, regulators and competitors will end up doing the same thing anyway." Self-disruption is a method of extending your own "post-death lifespan."
If the thing you're selling now — if being disrupted is inevitable — are you willing to be the disruptor yourself?

Philip Morris the man died in 1873 — he spent his entire life running that little shop on Bond Street, never knowing his surname had become the largest tobacco brand on earth. His company was bought by British partners, then later acquired by American investors. That name, "Philip Morris," after vanishing from storefronts for nearly half a century, was dredged back out of history by a group of Americans who had no connection to him, and hung on the New York Stock Exchange. No descendants inherited anything. Only a name floating in midair.

← Previous Contents Next →
Chapter 39 GE-Vernova

April 2, 2024. After 132 years — General Electric (GE) went from being one company to three. One of them is called GE Vernova — it inherited all of GE's energy businesses: gas turbines, wind turbines, grid equipment, hydropower, and nuclear. On its first day of independent trading on the New York Stock Exchange, its market cap exceeded $20 billion. This is a company "born as an adult" — it carries GE's century of accumulated power-generation technology and global installed base, now running afresh under an entirely new management team.

But the point of this story is not the day of the split — it is why "energy," inside GE, gradually went from being a primary engine to being an obscured gem, and how that value was released anew through separation.

In 1896, GE built the first commercial hydroelectric generator — beside Niagara Falls, converting water power into alternating current and transmitting it 26 miles to the city of Buffalo. In the 1900s, GE invented the steam turbine — and then, over the course of a century, GE's gas turbines, steam turbines, and nuclear technology became the backbone of global power generation. Roughly 30% of the world's electricity is generated by GE equipment or equipment built under GE-licensed technology.

But in the 2010s, GE's energy business was dragged down simultaneously by three things: first, the 2015 acquisition of Alstom's energy division for over $10 billion — a deal widely regarded as one of the most catastrophic acquisitions in GE's history (it bought in on the eve of a global gas-turbine market downturn). Second, renewable energy — wind power — had extremely thin profit margins but, driven by environmental policy, demanded continuous investment. Third, after 2008, GE Capital sucked away management's entire attention, and the energy business lost its original culture of technology-first prioritization.

GE Vernova's CEO, Scott Strazik, is an engineer who spent 30 years inside GE's energy and aviation divisions. At the time of the breakup, he said something chillingly blunt: "Before, we shared a balance sheet with the financial division — now we only use our own. Before, headquarters could kill one of our gas-turbine R&D projects because they needed to shift $500 million over to GE Capital — now that button isn't there anymore."

This is the essence of GE Vernova's "founding" — it is not a startup. It is an industrial republic that seceded from an empire. Its freedom is not "starting from zero" — it is "finally not having to be managed by someone else's logic."

Three things.
First: sometimes "separation" creates more value than "integration."
The rise in GE Vernova's stock price after independence reflects that investors assign a higher valuation to a "pure-play energy business" than to an "energy business bundled together with finance and healthcare."
In your company — is there a business unit that, standing alone, would be worth more than it is when lumped together with everything else?
Second: the scars of a bad acquisition can linger for decades.
GE's acquisition of Alstom was a balance-sheet destruction — it bought into a market that was being squeezed by the global renewable-energy transition. Not because the technology was poor — but because the timing was entirely wrong.
The last time you had an "acquisition impulse" — did you check what that market would look like five years later?
Third: separation is not abandonment — it is liberation.
GE's choice to spin Vernova into an independent company was not giving up on energy — it was admitting that "the optimal decision-making logic for three industries (aviation, healthcare, energy) is completely different, and having them commanded by the same capital-allocation committee will cause each to harm the others."
Do you have a division — whose poor performance is not because the team is weak, but because it is fundamentally incompatible with the native logic of your management?

The name Vernova comes from Latin — "verde" (green) + "nova" (new). But the revenue from its legacy gas-turbine division, selling old-style combined-cycle equipment, still accounts for the majority of its income today. It is a hybrid — "the power plants of the old world + the promise of new energy." Its future depends on whether this alchemy can survive the long transitional corridor of the clean-energy transformation.

← Previous Contents Next →
Chapter 40 Goldman-Sachs

1869, a German Jewish immigrant hung a brass plaque in a basement on Pine Street in New York: Marcus Goldman — Commercial Paper Trading. He was 48 that year, with nothing in his pocket but a leather billfold and an address book. Every morning, he took a stagecoach upstate to the jewelry merchants and leather dealers of New York, bought up their not-yet-due accounts receivable with cash — then returned to Manhattan the same day and sold them to the banks. He earned a thin spread in between. Today, that brass plaque bears another name — Goldman Sachs. Market cap: $274 billion.

Not financial engineering. Not trading algorithms. Not global expansion. Not capital engineering.
Just a horse, a ledger of names, and a leather billfold stuffed with promissory notes. Marcus Goldman spent his entire life doing only one thing: making strangers trust each other. He turned the "spoken credit" between merchants into a piece of paper that could circulate across the entire state of New York.
Then his son-in-law Samuel Sachs carried it into the world of investment banking. And then Goldman Sachs spent 157 years becoming the most respected — and the most feared — financial machine on earth.
This story has nothing to do with garages and code. It is about — the scaling of trust.

Marcus Goldman was born in 1821 in Bavaria, Germany. At 27, he came to America with the wave of German Jewish immigration. He spent over a decade as a small-time peddler — clothing, sundries. Then he noticed something: when upstate New York merchants sold goods to the big distributors in New York City, they were all issued promissory notes with "payment in 60 days." Those merchants needed cash every day to buy raw materials — but all they had in hand were promissory notes. Banks were unwilling to discount those "out-of-town, unknown" commercial papers — because there was no credit rating, and it was too labor-intensive.
Goldman set out every morning by stagecoach, walking through the shops of every town. He knew each shopkeeper's wife's maiden name, how many kids they had in school. He didn't need a credit score — he used his own two legs as the credit score.
In the afternoon, he returned to Manhattan and sold the promissory notes to the big banks at a slightly higher price — the banks were willing to take his paper, because "a note endorsed by Goldman had never once bounced."
He was not a banker. He was a "credit hauler" — at a time when no one was willing to verify credit, he used his own feet and his own time to verify it, and then re-sold that confirmed credit.
By 1882, his son-in-law Samuel Sachs had joined — and the firm was renamed Goldman, Sachs & Co.
Samuel brought something different to the table. What he saw was: commercial paper could only earn microscopic interest spreads. The truly big business was helping companies issue stocks and bonds. In the 1890s, Goldman began underwriting securities. Their first underwriting client was the United Cigar Manufacturers Company — a company that earned its living from the working man's five-cent coin.
And then Samuel made a decision that would distinguish Goldman from every competitor ever after: help small companies finance and go public. At the time, the House of Morgan only served the railroad and steel giants. Goldman was willing to underwrite mid-sized, growth-stage retailers and manufacturers — because Samuel believed "mid-sized does not mean low-quality; it means still growing."
In 1906, they took Sears, Roebuck — the then-surging department-store mail-order company — public. Then Woolworth. Then Continental Can. Then Merck.
This strategy was inherited in spirit by every subsequent generation of Goldman partners — find the companies not yet large enough to be coveted by Morgan, and then ride alongside them as they grow into giants.

And then what truly turned Goldman into a god-tier name was Sidney Weinberg.
Weinberg came to Goldman at 16 as an assistant to the janitor — mopping floors, wiping inkwells, delivering packages. He later helmed Goldman for 40 years — and was called "Mr. Wall Street."
With his acne-scarred face and thick Brooklyn accent — Weinberg became the most trusted corporate advisor in all of America. Every day, he was the first to arrive at the office and the last to leave. He forbade any Goldman partner from holding personal investments outside the firm — "If you believe what we do for others is right — then you put all your own money in the same things."
After the Great Crash of 1929, the whole of Wall Street was reduced to rubble. Goldman itself nearly went bankrupt — because one of its investment trusts went to near-zero in the crash. Weinberg swallowed the humiliation and spent a full decade rebuilding the firm's reputation — retrieving clients one by one from the wreckage.
1956, the Ford Motor Company was going public — at the time, the largest IPO in American history. When old Henry Ford's grandson chose the underwriter — he chose Goldman Sachs. Not because Goldman was the biggest. Because Weinberg had spent nearly ten years building personal trust with the Ford family. And one of Weinberg's core principles was — never write a letter recommending an IPO. He simply stayed in the living rooms of Detroit, helping the Ford family untangle their family shareholding structure day after day — until one day, they themselves said, "It's time to go public."

Now stop for a moment.
Do you know who your most important potential client is — the one who might change your destiny — and how many years have you invested in them, "not selling anything, just helping"?
Weinberg invested ten years in Ford before landing a single mandate. In those ten years, he sold nothing. He just showed up every day. Can you do that?

Three things.
First: a trust-hauler is harder to replicate than a technology-hauler.
Goldman's sole core competency over 157 years — is not algorithms, not models, not trading speed. It is "the trust network." Governments, corporations, billionaires — they hand their largest transactions to Goldman because they know Goldman will never let a single word leak after the deal closes.
Do your customers choose you because "trusting you is safer than trusting anyone else" — or because your features are marginally better than the next guy's?
Second: don't chase big clients — raise the clients who will become big.
In its early days, Goldman didn't compete with Morgan for railroads. It went to the retailers, the manufacturers, the consumer companies that were "beneath the standard" at the time and helped them finance. Fifty years later, consumer retail had become the largest sector. Grow with the client.
Among your current clients — is there one that's "still small now but you are certain won't be small ten years from now"?
Third: always express confidence with your own money.
Sidney Weinberg banned partners from making outside investments — their entire net worth had to be placed in the assets Goldman itself underwrote. This looked like a restriction. In reality, it was the brand: for every project Goldman pitched — the person recommending it had already bet his life on it.
Do you dare require every one of your sales directors — to hold your own company's stock as their entire net worth — without being able to sell a single share of anything else?

The emblem of Goldman Sachs is a bull — in the lobby of 1 Wall Street. One hundred and fifty-seven years. It has never fallen. Not because it was the biggest. Because every time it faced a choice — pick long-term reputation, or pick short-term profit — its partners, generation after generation, almost always chose the former.
The first promissory note inside Marcus Goldman's leather billfold — the name has been rubbed illegible by time. But that man lived into his nineties, and from the window on Pine Street, he could always watch the entire financial district rise up around him.

← Previous Contents Next →
Chapter 41 Texas-Instruments

Summer 1958, Dallas. Every employee at Texas Instruments was on a mandatory two-week vacation. Except for one new hire named Jack Kilby — he'd only been there three months, not long enough to qualify. Alone in the empty lab, he did one thing: he etched transistors, resistors, and capacitors onto a single slab of germanium crystal. The world's first integrated circuit was born — because a lonely engineer had nowhere else to go.

Not a team effort. Not a massive budget. Not a national project. Not market demand.
Just him, in a deserted building, using one piece of material and a few idle machines, pulling the electronics industry out of a world of hand-soldered wires and into a world of microscopic silicon printing. More than half a century later, Texas Instruments has a market cap of $262.8 billion.
Let me tell Jack Kilby's story first. After that, we'll talk about why the most important innovations are often not born inside a pressure cooker — but when no one is watching.

Jack Kilby. From Kansas. A big man who spoke slowly, like a small-town general store owner. He graduated from the University of Illinois with a degree in electrical engineering in 1947. When he joined Texas Instruments in May 1958, TI had a company-wide mandatory vacation policy — two fixed weeks every summer when everyone had to be off at the same time, to reduce the risk of anyone tampering with individual workstations. Kilby had no vacation time — too new.
He sat alone in the lab, staring at the "tyranny of interconnects" — the most fundamental physical problem in the electronics industry at the time: every transistor, every resistor, and every capacitor had to be hand-soldered with metal wires onto a printed circuit board. Engineers worked like embroiderers, soldering hundreds of wires under a magnifying glass. When circuits grew complex enough to involve thousands of transistors, this manual method became physically impossible. The U.S. military had already given TI a "micromodule" research contract, but that was about soldering together standardized small square circuit boards — still wiring, at its core.
Kilby quietly turned over a single thought — what if all the components could be made from the same semiconductor material? No metal interconnect wires needed — because the material itself could form resistance.
He sliced a germanium wafer into thin strips, used etching and impurity diffusion to create several transistors on it, and used the conductive properties of the same material to form resistors. Then he connected ultra-fine gold wires from the devices to external pins — the thing looked like a little glass chip covered in golden spiderwebs, ugly as hell. He hooked it up to a power supply and an oscilloscope — a continuous sine wave oscillation appeared on the screen, the waveform intact.
Later generations would do integration on silicon — but what Kilby proved that day on germanium was that "everything can be done inside a single piece of material." It was the first integrated circuit in human history.
Almost simultaneously — in California, Robert Noyce at Fairchild Semiconductor independently conceived of an integrated circuit using vapor-deposited aluminum layers on a silicon plane as interconnects — a design much closer to today's CMOS processes. In 2000, Kilby received the Nobel Prize in Physics — Noyce had already passed away. In his acceptance speech, Kilby made a point of mentioning Noyce: "If he were still alive, he would be standing here. We each stood at one end."
The integrated circuit changed human history — from computers to mobile phones to pacemakers to GPS to AI chips, the physical foundation of everything traces back to that empty TI lab in the summer of 1958.

And then Texas Instruments became a strange company. It has its own chip fabs — but it also makes calculators, educational toys, military radar, and the core optical engine for projectors (the DLP chip). It doesn't focus solely on computing like Intel, or solely on communications like Qualcomm. It makes many different kinds of chips — analog, embedded, industrial — all the nerve endings that connect the "real physical world" to the "digital world."
This is TI's moat: not the fastest, but the most ubiquitous. Every car's brake sensor, every air conditioner's temperature controller, every factory robot — hidden inside each is an unmarked TI analog chip. You'll never see one in your lifetime, yet it appears beside you more frequently than any brand you know.

Three things.
First thing: Major breakthroughs often emerge in moments when no one is watching.
Kilby had no team pressure, no weekly reports, no project management software. Just an empty lab and two weeks of uninterrupted time.
When was the last time you had two consecutive weeks of undisturbed "empty lab time"?
Second thing: The "tyranny of interconnects" is the fundamental bottleneck in every industry.
The electronics industry's interconnect bottleneck was hand-soldered wires — solved by integration. In your industry — what is the equivalent of "hand-soldered wires"? What is the step that turns all scattered parts into a single closed system?
Third thing: Be the bridge between the digital world and the physical world.
TI's core isn't building the fastest CPU — it's building the chips that convert physical signals (temperature, pressure, speed, light) into digital signals, and then convert digital signals back to precisely control physical devices. Every time industrial automation levels up one notch — TI is the supplier of that bidirectional translation chip.
In your industry — can you become the translator connecting the "old physical world" and the "new digital world"?

Jack Kilby passed away in 2005. He never became a billionaire — TI's compensation system didn't give him a stock-option jackpot. A reporter asked him if he regretted not getting rich like Silicon Valley engineers. He said: "The things I made are now in every child's hearing aid, in every heart patient's pacemaker, in spacecraft and ocean probes." "That's more interesting than money."

← Previous Contents Next →
Chapter 42 RTX---雷神技术

1922, Cambridge. Three MIT engineers — Laurence Marshall, Vannevar Bush, and Charles G. Smith — founded Raytheon in an old warehouse. Its first product was not a missile — it was a household refrigerator compressor. But it didn't cool anything. They turned this failed compressor into an electronic component that could replace vacuum tubes — and it became the heart of the world's first commercial car radio.

Not defense. Not radar. Not missiles. Not satellites.
Just three friends doing refrigeration research at MIT — who discovered that their refrigerator compressor experiment was physically correct, just pointed in the wrong direction. They wanted to "cool things down" — and instead ended up creating a component that could precisely control the flow of electrons. That first failed "Raytheon Tube" spent the next decade and more inside the radios of every American car.
Then World War II broke out. Raytheon became synonymous with radar — seven out of every eight radar installations the British deployed over the English Channel had their magnetron cores manufactured by Raytheon in the United States. And then from radar came the microwave oven (yes — Raytheon engineer Percy Spencer discovered in 1945 that the magnetron could melt the chocolate bar in his pocket. The first microwave oven was called the Radarange, weighed 340 kilograms, and sold for $5,000).
In 2020, Raytheon merged with the aerospace division of United Technologies Corporation to become today's RTX, with a market cap of $239.3 billion. It is one of America's largest manufacturers of air-defense missiles, Patriot systems, and next-generation sensor networks.

Three things:
First thing: A refrigerator failure became a car radio — this is the classic byproduct of "one direction that's physically correct but commercially wrong." In the experimental scrap heap at your company — is there a potential secondary commercial direction hiding in the underlying physics?
Second thing: The magnetron was the secret core of radar in the 1940s — by 1945 it had become a popcorn heater. Military-to-civilian conversion doesn't have to be a "downgrade" — it can be a lateral transfer of physical function: microwaves went from detecting enemy aircraft to vibrating water molecules into heat.
Third thing: Merger and integration capability — Raytheon used UTC's engines and aerospace units to complete its own leap from "radar and missiles" to a full aviation-defense-propulsion system. The last time you merged — did you put both supply chains into the same mathematical system?

← Previous Contents Next →
Chapter 43 KLA---科磊

1997, an MIT-trained engineer named Ken Levy became CEO of a new company formed from the merger of KLA and Tencor. He ran an "invisible company" — its equipment doesn't manufacture any chips, but it inspects every defect in the chip manufacturing process that the human eye cannot see. If the lithography machines at TSMC and Intel are the paintbrush — KLA's machines are the microscope that examines every brushstroke. Today KLA Corporation has a market cap of $229 billion. It has never produced a single chip — but every advanced chip on Earth must pass through KLA's wafer inspection and metrology equipment before it can leave the cleanroom.

Not chip manufacturing. Not lithography machines. Not design software. Not materials breakthroughs.
It's "yield management" — catching invisible nanometer-scale defects in the fab after every process step using optics, electron beams, and lasers. When KLA was founded in 1976, yield improvement relied entirely on engineers eyeballing wafers under optical microscopes. Ken Levy turned defect detection into automated statistical process control — and then, with every nanometer the semiconductor industry shrank, it needed more sensitive inspection. Now, in 3nm processes, KLA's equipment can identify defects smaller than a virus and sentence a wafer to death — before it ever becomes scrap.

Three things:
First thing: Don't be the lead actor — be the referee that no one can do without. Chip fabs can switch lithography suppliers (ASML, Nikon, Canon), but once defect detection and yield calibration have been written into the entire fab process flow by KLA — replacing it means shutting down the factory and starting over.
Second thing: Every time the process node shrinks — inspection difficulty rises exponentially, and with it the barrier to entry. This is not a race for "first place" — it's a race for "only place."
Third thing: Establish authority at the invisible level — when defects are at the 2-nanometer scale, the customer has no choice but to trust the inspection equipment's results. KLA's output is the wafer fab's truth.

← Previous Contents Next →
Chapter 44 Wells-Fargo

1852, two Eastern merchants stepped off a boat onto the San Francisco docks. The entire city was in a gold rush — everyone was in the mountains digging. They didn't go to the mountains. They set up a table on the dock with a sign that read: Leave your gold dust with me — it will still be here tomorrow.

Not digging for gold. Not selling shovels. Not running hotels. Not escorting gold shipments.
Just Henry Wells and William Fargo. In the midst of the most frenzied gold rush, they did something considered utterly unglamorous at the time — they stored gold dust for strangers, issued money orders, and moved money from California back to families in the East.
Only a tiny fraction of gold rushers ever found real gold. But every single person who deposited gold dust with Wells Fargo — was profit for the bank. Wells Fargo's market cap today is $225 billion. The stagecoach is still printed on its logo, still running.
Let me tell its story first. After that, we'll talk about why the cool-headedness of "don't chase the gold mine, just run the bank" — wins in every mania.

The California Gold Rush erupted in 1849. Hundreds of thousands of people poured into Sacramento and San Francisco from the American East, from Europe, from China. Gold in the silt — people went so mad they hollowed out toothbrush handles to stuff with gold dust.
Henry Wells and William Fargo — both co-founders of American Express — saw something different in 1852. Those gold rushers needed to ship gold dust and silver ingots to their families back East, or needed a safe place to store their gold dust. At the time, California had extremely few banks — and most had no national agency network. Gold dust deposited in Sacramento couldn't be withdrawn in New York. And the stagecoach and steamboat routes were crawling with bandits and bad roads.
They opened their first office on Montgomery Street in San Francisco. Then they bought a fleet of stagecoaches and a team of armed guards. What they did was "transport gold dust + issue money orders + store gold bullion" — three unsexy basic banking operations. Then, with San Francisco as headquarters, they spread inland station by station along the gold rush routes — Sacramento, Stockton, every noisy encampment in the mining districts.
While others sold jeans and pickaxes, Wells Fargo took in gold dust, assayed its purity, and issued money orders redeemable on the East Coast — collecting a small commission each time. Every grain of gold dust that traveled two hundred miles without being stolen — pressed its brand reputation one layer deeper.
By 1855, over 50 "bank and express" companies in California had collapsed. Wells Fargo didn't. Because it spent more on vaults in every gold-mining town and armed escort on every stagecoach than any competitor.

In 1905, Wells Fargo was split up — the banking and express businesses separated. The express business evolved into a UPS-scale logistics operation, and the banking business was sold to what would later become the Norwest Group.
But here's what truly takes your breath away — when Norwest acquired the original Wells Fargo in 1998, it chose to keep the name "Wells Fargo" instead of its own. Because the brand of the Western stagecoach was worth far more than the registered name Norwest.
Then came the 2008 financial crisis. Wells Fargo did what all banks were doing — took bad debts, wrote down assets — but unlike Citigroup and AIG, it didn't need a government bailout. Why? Because its core business was too old-school — deposits + loans + mortgages for middle-class families. No massive exposure to investment banking gambles. Warren Buffett was at one time one of Wells Fargo's largest shareholders — he bought Wells Fargo his whole life because "the business it does, I understand — people put money in, and it lends it to others to buy homes."
Then in 2016, the fake accounts scandal erupted. Wells Fargo employees, driven by insanely aggressive account-opening KPIs, opened millions of fake accounts without customers' knowledge. CEO John Stumpf was grilled at Senate hearings and resigned. The Federal Reserve imposed an asset cap sanction on Wells Fargo — which remains (as of 2025) not yet fully lifted.
This is the darkest page in the stagecoach story. A bank that spent 160 years guarding "trust" through gold rushes, earthquakes, and the Great Depression — had the very thing it was supposed to guard most corroded by its own internal sales culture.

Stop here for a moment.
Has your growth North Star — traveled so far that it has crushed the very foundation stone that first brought you into existence? Wells Fargo's sign at the gold-rush camp — "your gold dust will still be here tomorrow" — took 160 years to build, and 3 years of sales KPIs to destroy.
Which of your KPIs is currently locked in mutual slaughter with your core mission?

Three things.
First thing: In the most feverish speculative frenzy, sell the most level-headed service.
When everyone was digging for gold — Wells Fargo was storing it. When everyone was chasing NFTs and crypto — Wells Fargo was still doing home mortgages. Cool-headedness isn't sexy. Then the frenzy recedes, and the cool-headed are still standing.
What is your industry chasing right now? Are you the one person unwilling to chase it — and if so, will your cool-headedness become your greatest future premium?
Second thing: The stagecoach isn't nostalgia — it's a promise.
Wells Fargo's stagecoach trademark isn't decoration. It keeps reminding every employee — this company existed before railroads and the internet existed. That kind of historical density cannot be replicated by any new bank.
Does your brand carry a kind of temporal gravity that makes employees feel "I am inside something important that is far older than I am"?
Third thing: A bank's real asset isn't money — it's the ledger.
In the 19th century, Wells Fargo was a "mobile ledger" — physically recording the owner of every grain of gold dust. Today it is one of the largest custodians and record-keepers of assets under management globally. And a ledger — once everyone trusts you — you become the default clearing layer for all transactions. And then you cannot be replaced, because you have become everyone's shared memory.
Is your business — becoming the "undeletable ledger" of an industry?

From the Gold Rush to the fake accounts scandal — Wells Fargo learned this: trust, that thing — takes 160 years to build, and only one KPI compromise in an internal meeting to destroy.
The stagecoach is still running — it's just that the strongbox inside its carriage no longer holds gold dust, but digits. And whether it can run for another 170 years — depends on whether it still remembers why in 1852 it chose to load that stagecoach with gold dust instead of gold ore.

← Previous Contents Next →
Chapter 45 QUALCOMM

1985, San Diego. Seven engineers started a company in a garage. Their first product was a vehicle-mounted satellite positioning terminal — big and expensive. No customers bought it. Then they did the reverse: they stopped selling terminals. They sold the "math inside the head." They took a military spread-spectrum technology called CDMA and turned it into the communications standard for every mobile phone on the planet. Today, Qualcomm has a market cap of $222.5 billion.

Not making phones. Not making base stations. Not making chips. Not making operating systems.
Just Irwin Jacobs. MIT PhD, former USC professor. His story isn't "couldn't find customers, so he died" — it's "first product failed, abandoned hardware, and used mathematical formulas to rule the entire mobile communications industry."
Qualcomm doesn't manufacture any consumer products. Its chips (Snapdragon) and communications patents are the indispensable foundational technology inside every smartphone. Every cent you pay when buying a phone — a small fraction of it becomes a Qualcomm patent royalty, flowing into a massive patent portfolio in San Diego.
Let me tell this story first. After that, we'll talk about why "abandoning hardware, going all-in on software and standards" is one of the most profitable pivots in the history of tech.

Irwin Jacobs was already well-known in both academia and industry before founding Qualcomm. He completed his PhD dissertation on information theory at MIT, then founded the communications research group at UCSD. Through U.S. military-funded research, he was exposed to a spread-spectrum communications technology called CDMA — originally used for military encrypted communications, because spread-spectrum signals are extremely difficult to detect and jam.
In 1985, Jacobs and Andrew Viterbi (inventor of the Viterbi algorithm) co-founded QUALity COMMunications — Qualcomm. The first product was called OmniTRACS — a satellite communications terminal mounted on trucks. No trucking company wanted to buy it.
Then Jacobs made a decision: bring CDMA from military use into civilian cellular networks. At the time, the entire communications industry had already invested hundreds of billions of dollars in TDMA (time-division multiple access) — the GSM standard in Europe, TDMA in the United States. Everyone said CDMA was physically impossible in a mobile environment — because the power control between mobile stations and base stations was far too complex.
In 1991, Qualcomm built an experimental CDMA network in San Diego with its own hands. They invited carriers from around the world to come see — in that demonstration, the same frequency supported 10 to 20 times more call capacity than TDMA. Carrier inboxes exploded.
Then Qualcomm made an institutional decision that would later change its destiny — it sold its phone manufacturing and base station manufacturing divisions to Kyocera and Ericsson, turning itself into a pure "chip design + patent licensing" company.
It no longer manufactured anything. It only collected the "Qualcomm tax" — every 3G, 4G, and 5G phone, no matter who made it — paid Qualcomm a percentage of the device's selling price as a patent licensing fee. Apple, Samsung, Huawei — everyone pays.
This was a nuclear-bomb-scale pivot: cutting away your own product lines and keeping only the core standard.

But this "Qualcomm tax" triggered a global regulatory storm. China's NDRC fined Qualcomm $975 million for antitrust violations in 2015. The Korea Fair Trade Commission, the European Commission, and the U.S. FTC all launched successive antitrust lawsuits against Qualcomm's "no license, no chips" bundling model. Apple and Qualcomm fought a global patent war for three years — then suddenly settled in 2019, with Apple signing a six-year licensing agreement.
Qualcomm won this round — because it holds core patents on key technologies within the 5G standard. And 5G is the infrastructure of the 21st century. When the entire world began building 5G networks — no matter how many lawsuits Qualcomm won or how many fines it paid — its payment window would never close, because the entire industry has no alternative path. Every phone maker must buy Snapdragon chips from Qualcomm or obtain licenses for CDMA/5G standard-essential patents — or both.

Three things.
First thing: The failure of the first product is not the failure of the company — it's the funeral of that specific product.
No one bought the OmniTRACS satellite terminal. Irwin Jacobs didn't cling to it — he took OmniTRACS's underlying technology, turned it back into raw components, and reconstructed it into a CDMA cellular system. Those are two completely different markets.
If your first product dies — do you dare to extract its internal recipe and feed it into the next, entirely different product?
Second thing: Don't make things — make the "standard."
Qualcomm went from manufacturing phones to "only designing chips + setting communications standards." A standard isn't a product — a standard is "a toll gate this industry must pass through." If you hold standard-essential patents — others aren't competing with you, they're using your license to legally survive.
In your industry's standards — are your patents or rules hidden inside?
Third thing: Patents aren't a protective shield — they're a perpetual right to collect.
Qualcomm spends over $8 billion a year on R&D — most of it on standard-essential patents for the next generation of communications standards. Then those patents become licensing fees that can be collected for decades — not just from this year's phones, but from every new phone in every future year.
On what kind of asset — can you do R&D once, and collect money for a lifetime?

Irwin Jacobs is now over 90 years old. He handed the CEO role to his son Paul Jacobs, who then passed it to Steve Mollenkopf, and now it's Cristiano Amon. In his lifetime, Qualcomm went from a truck satellite company to the "foundational technology overlord" to whom every phone in every pocket on Earth owes a small fee.

← Previous Contents Next →
Chapter 46 American-Express

1850, three stagecoaches set off from Buffalo, New York. They weren't carrying mail — they were carrying money. Inside the strongboxes on board was cash and gold, sealed with wax. Hanging from the front of the stagecoaches was a hand-painted wooden sign — "American Express." Today, its market cap is $213 billion.

Not a bank. Not an investment bank. Not the invention of the credit card.
Just American Express. It began as a stagecoach express company — like something out of a Western — hauling cash and parcels between small towns in New York State. Then, through every generation of communications revolution — telegraph, telephone, internet — it kept on "shipping trust." It turned a purple plastic card into a shared payment faith between the global elite and small and medium-sized merchants alike.
Let me tell its story first. After that, we'll talk about why "premium faith" is a moat harder to copy than any technology.

In 1850, three fiercely competing Eastern stagecoach express companies merged into one — American Express. Why merge? Not because they liked one another. It was because the stagecoach era was extremely inefficient — different route companies refused to handle each other's parcels, and packages could be tossed on the ground at relay stations.
In the first year after the merger, they unified cash and gold express service from New York to Buffalo under a single operation — cutting the delivery time in half. Then the railroads came. Stagecoaches were replaced by trains. Amex didn't stubbornly keep feeding horses. It swapped the stagecoaches for strongboxes inside train cars and kept hauling cash.

1891 — the critical turning point. Amex invented the traveler's cheque.
Not new technology — a stroke of genius in trust mechanisms. You deposited cash with Amex, and they gave you a booklet of pre-printed denominations. You traveled to Europe, found any bank, signed your name twice — and collected cash. Lost it? Canceled and refunded immediately. Before credit cards and electronic payments existed, this was one of the greatest travel payment tools in human history.
In 1914, World War I broke out, and masses of American tourists found themselves stranded in Europe. Amex undertook the largest manual labor of trust in its history up to that point: they sent people — one by one, bank by bank, hotel by hotel — to persuade them to accept their traveler's cheques. Not documents. Human beings walking to every front desk and saying: "Accept this piece of paper — if there's any problem with it, I will cover the loss."
That campaign turned Amex into, in the eyes of every hotel manager in Europe, "that American who will never let you lose money."

1958 — Amex issued its first charge card. Purple.
The travel gene seeped into the product's color — it wasn't the generic "swipe it anywhere" payment card that Visa and Mastercard were. It was "if you want to be respected at the finest restaurants and the most exclusive lounges — you need a purple card."
This isn't a technical barrier. It's an emotional threshold.
In the 1970s, Amex transitioned from charge cards to revolving credit cards — but kept its elite pricing of "balance must be paid in full or in large installments." It charges merchants higher fees than Visa and Mastercard — not because its technology is better — but because the Amex cardholders it brings spend, on average, five times more than others. High-end restaurants have no choice but to accept it.

Then the 2008 financial crisis. Amex is not just a payment network — it is also the card issuer. It lends its own money to cardholders. When the subprime bubble burst, unemployment soared, and large numbers of Amex cardholders began defaulting — Amex became one of the most heavily shorted financial institutions at the time. Analysts warned its bad debts would drown it.
CEO Kenneth Chenault made a counter-consensus decision — he didn't tighten credit immediately. Instead, he carried out gentle restructurings for some customers in distress, while slashing costs aggressively and stabilizing the merchant side. Then, as the crisis receded in 2010, Amex's bad debt rate recovered far faster than that of any commercial bank.
From 2021 to 2025, Amex began facing heavy competition from tech payment companies — but its strategy didn't change: don't fight Stripe and Square in a price war over the long tail of small merchants. It focuses on "business travel + premium spending + the younger generation's blended demand for experiences and status." The "invitation-only tiering" of the Platinum Card and the Centurion Black Card has become one of the most successful membership mechanisms in human history.

Stop here for a moment.
Do you have a paid product — not because it's cheap or feature-packed — but because "acquiring it in itself makes you feel like you've entered a club"?
Amex isn't selling payments. It's selling a psychological certification that says "I belong to a certain tier." In your industry — how much is this "psychological certification" worth?

Three things.
First thing: From stagecoaches to AI — core technology can die generation after generation, but the trust brand must not die.
Amex has survived stagecoaches, trains, telegraphs, telephones, the internet, blockchain — every physical method of "sending money" has changed. But Amex has always been "sending trust." Its operational details have been completely reinvented every decade. Its brand promise has never changed.
Your brand — strip away all technology — what is the one thing that "never changes"?
Second thing: Reject the long tail — defend only the highest per-customer spending.
Amex's merchant fees are the highest, and its customer base is the narrowest. But it doesn't fight Stripe and Square — because it doesn't need to. The thin top layer of global business travel and premium spending — is enough to sustain it.
Is your greatest profit — scattered across too many low-value customers? Do you dare to proactively hand your bottom-tier customers to competitors — and focus on serving the top of the pyramid?
Third thing: "Invitation-only" creates more desire than advertising.
The Amex Black Card has never been publicly promoted. You can only be invited. This created the cheapest yet most valuable marketing in human history — because the curious people who are rejected turn into unpaid brand ambassadors.
Can your highest-tier product line — go from "anyone with money can buy it" to "only those we choose can enter"?

One hundred and seventy-six years have passed since American Express set off on that stagecoach. It no longer ships gold. It ships trust. And trust — in the world of commerce — is more precious than gold.

← Previous Contents Next →
Chapter 47 Citigroup

1986, a stockbroker born in Brooklyn sat in his Manhattan office and hung on the wall an organizational chart he'd drawn in crayon himself: at the top was "Us" — a consumer loan company so small it was negligible. Below it was a string of names he wanted to devour: American Express, Aetna, Citicorp — "this list is what we're going to buy in the future." His business partner tore the chart off the wall and said, "Don't let anyone see this. People will think you've gone crazy enough to be locked up." Then he spent 12 years eating every name on the list, one by one — including the largest bank in America. Today Citigroup has a market cap of $211.8 billion.

Not financial engineering. Not Wall Street rules. Not regulatory arbitrage. Not insider trading.
Just Sandy Weill. A poor Jewish kid from Brooklyn who dropped out of NYU after a year and a half and sold stock ticker machines in lower Manhattan. He never invented a single financial product in his entire life. He just did one thing that only one person in the entire history of finance has pulled off — he acquired companies bigger than his own, again and again. Each time, using share swaps to swallow them. Each time, once the swallowing was done, he deleted the opponent's name — and replaced it with his own.
In 1998, he completed the boldest merger in financial history — Travelers Group merged with Citibank to become Citigroup. That merger shattered the Glass-Steagall Act of 1933 head-on — before it was even formally repealed.
Let me tell his story first. After that, we'll talk about why "one man's desire to win" can simultaneously be the most powerful engine in history — and the spark that nearly burned down the entire financial machine.

Sandy Weill was born in 1933 in Brooklyn, New York, to a family of Polish Jewish immigrants. He talked fast, walked fast, and his brain moved faster than his mouth. His first job after graduation was selling stock ticker machines — those old-fashioned narrow-paper-tape stock-quote machines. He walked into buildings one by one, starting from the southernmost tip of Manhattan, pressed the elevator button, knocked on doors — and learned one thing: the more doors you knock on, the more orders you get. No ceiling.
In 1960, he and three friends founded Carter, Berlind, Potoma & Weill with very little capital — a tiny brokerage firm. For the decades that followed, his process was completely consistent: first, run his own company to extreme profitability — then use the stock premium to buy a company larger than his own, but poorly managed. After the purchase, clear out the entire management of the target and implant his small core team into the opponent's offices. Then profits doubled, the stock price rose, and he bought the next one, bigger still.
Each "small-swallows-large" operation was like a snake swallowing a cow — Wall Street said every single time, "This time he's definitely going to choke." Then he'd swallow it and digest it.
His acquisition list included — Hayden Stone (1970), Shearson (1974), Loeb Rhoades (1979), American Express (in 1981, he briefly sold his company to Amex and became its president, then was forced out by a brutal internal power struggle — this was the biggest failure of his career). Then in 1986, he took over a tiny, all-but-dead consumer loan company called Commercial Credit — he got it back to profitability in just one year, then used its stock to acquire Primerica, then Smith Barney brokerage, then Aetna's insurance brokerage, and then, in 1998, the merger of Travelers Group with Citibank — giving birth to the largest financial services company in the world, Citigroup. That year, the Glass-Steagall Act was still in effect — "commercial banking" and "investment banking" were not supposed to belong to the same company. But the reality of the merger and the Federal Reserve's grace window directly triggered the act's complete repeal in 1999.
Then the 2008 financial crisis. Citi was one of the core distributors and holders of subprime mortgages, SIVs (structured investment vehicles), and CDOs (collateralized debt obligations). After the crisis erupted, Citi's stock price plunged over 95%. The U.S. government injected $45 billion to prevent a global-scale collapse — because Citi had operations in over 100 countries at the time, and liquidating it was "too big to fail" — no longer a debate, but a panic. Sandy Weill said something in a television interview that made the whole world smile ruefully: "I think my biggest mistake was — letting that wall be knocked down."

Stop here for a moment.
What did it feel like — the last time you were struck by the words "my biggest mistake" coming from the mouth of one of the most successful people in business history? Weill spent a lifetime knocking down walls — then admitted the wall should not have fallen. This is a strange truth: the thing he wanted to win most — "go all the way" — was precisely the variable the system ultimately couldn't bear.
Do you have a deepest desire — that, if you obtained it, might destroy you instead?

Three things.
First thing: Acquiring a company bigger than your own — is possible.
Weill's entire methodology was "seize the profit gap of a poorly managed giant." If your managed profit margin is 30% and theirs is only 5% — your stock premium will naturally cover the target's book value post-acquisition.
Do you currently have this profit-margin gap between yourself and a larger company in your industry that's more poorly managed than you?
Second thing: Every time you lose your position — start again from somewhere smaller.
Weill was 52 when he was forced out of Amex. He could have retired. He took over a micro consumer loan company that everyone thought was beyond saving. Then from that foundation, he paved the way to Citi, which became far bigger than Amex.
If you were pushed down from the highest position right now — do you have a "little brick house closest to the ground" where you could start laying walls again?
Third thing: He proved — one person's ambition can rewrite an entire nation's financial laws.
Do not underestimate the power of "one individual's will to smash through a wall" — the Glass-Steagall wall was shattered by Weill's personal acquisition logic and merger negotiations. And that same quality is what made the system fragile later. One person's will was right — and then that same thing, at a different point in time, became destruction.
Do you have a "rule you very much want to knock down" — and what would the 20 years after knocking it down look like?

After retiring, Sandy Weill didn't stay on Wall Street. He founded a small non-profit education foundation that trains public school principals — he calls it "the exact opposite of acquiring companies — here you can never think about quarterly reports."

← Previous Contents Next →
Chapter 48 T-Mobile-US

1999, the CEO of Deutsche Telekom made a decision that had every boardroom expert shaking their head — spend $35 billion to buy a tiny American operator called VoiceStream. At the time, the company had fewer than 2 million subscribers, and its network coverage was riddled with holes like Swiss cheese. They said he was paying the "fool's tuition" to enter the U.S. market. Today T-Mobile US has a market cap of $209.7 billion. It merged with Sprint and is now the second-largest wireless carrier in the United States.

Not a technological revolution. Not a spectrum advantage. Not capital overwhelming the field.
Just John Legere — who took over as CEO in 2012. The first thing he did was not reorganize the balance sheet. He took off the suit, put on a bright pink T-shirt, wore a baseball cap, and started directly cursing out AT&T and Verizon on his own Twitter — with profanity, with provocational videos, with live-streaming himself eating McDonald's. Then he got rid of "contracts" — that two-year binding clause that American consumers hated more than anything. He said: "We don't need to chain you to a wall to keep you here. If we build a good network and offer better prices — you won't leave."
This was called the Un-carrier strategy. The entire industry thought he was a lunatic destroying his own profits. Then T-Mobile went from fourth place all the way to second. Legere handed the CEO baton to Mike Sievert in 2020. At the handover, he said one thing: "We don't follow the rules anymore — we rewrite the rules every six months."

Three things:
First thing: Remove the "norm" the industry hates most — it becomes the only reason consumers need to download your app.
Second thing: Let the CEO himself be the social media battle axe — Legere didn't outsource social media; his very person became the brand's rebel persona. Do you dare let your CEO become the brand weapon that "breathes fire at competitors every single day"?
Third thing: Don't bind users — and that binds them the most. "You don't have to stay" is an extreme form of confidence — and consumers can sense that confidence intuitively.

← Previous Contents Next →
Chapter 49 SanDisk---闪迪

1988, a 65-year-old Israeli immigrant founded a company called SunDisk in California. His hair was completely white. His accent was thick. He had already failed twice before founding this company. Investors said to him: "You're starting a company at retirement age?" He smiled and nodded. Today SanDisk has a market cap of $208.7 billion. The flash memory inside every smartphone, every camera, every drone — is tied to this man's inventions.

Not a semiconductor giant. Not a Samsung-style full supply chain. Not government-backed. Not a young Silicon Valley myth.
Just Eli Harari. In the 1970s, he worked on electro-optical systems for the Israeli military, then went to the United States to work at Honeywell and Intel — in Intel's memory division, he caught the first glimmers of the floating-gate transistor (EEPROM). But his boss said: this direction is not Intel's core.
In 1988, he resigned and founded SunDisk in a small office in Santa Clara. He bet everything on a new-concept "non-volatile solid-state storage" — NAND flash memory. Data stored inside stays there even when the power is off. No mechanical arm needed, no magnetic head — a pure silicon chip.
For the first three years, no major customer placed a large order. His first big order came from Fujitsu, for internal storage in digital cameras. Then Kodak and Canon followed. Then Sony and Nintendo. Then the iPod arrived, bringing the "Shuffle" with it — it used flash memory chips. Then came phone cameras, 4K video, laptop SSDs, data centers, and enterprise SSD arrays.
In 2016, Western Digital acquired SanDisk for $19 billion — but the SanDisk brand remains to this day. After selling the company, Harari did something very strange — he poured all the money he had made into a startup working on solar methanol. He said: what flash memory does is trap electric charge in silicon, something you can do for a lifetime; now I want to use sunlight to fix carbon into liquid fuel — same logic, just a different element.

Three things:
First thing: Start a company at retirement age, carrying the experience of two failures — and the concept of "flash memory" needed exactly this kind of long-term madman who didn't care at all about short-term profits.
Second thing: Data doesn't vanish when the power goes off — this extremely simple physical principle became a market generating tens of billions of dollars in annual revenue.
Third thing: Toshiba invented NAND, Intel didn't bet on it — and the market was handed to a sixty-something immigrant. A big company's "not our core" is, time and again, someone else's life's core.

← Previous Contents Next →
Chapter 50 IBM

1914, a 40-year-old tabulating-machine salesman walked into a company called CTR. He saw everyone doing their own thing — the tabulating machine division, the time-clock division, the scale division — three piles of machines that didn't speak to each other. He did something no one else had done: he borrowed a massive sum, signed a contract heavily stacked against himself — and tied his entire personal fortune to this company. Then he said: from today onward, we do only one thing — "Think."

Not inventing the computer. Not writing code. Not the semiconductor revolution.
Just Thomas Watson Sr. A man who had been a traveling salesman and was fired by his previous employer for being "too aggressive in style." He took a company that made a hodgepodge of machines and renamed it International Business Machines — IBM.
Then he and his son, Thomas Watson Jr., survived every technological inflection point of the 20th century — tabulating machines to mainframes to personal computers to IT services to hybrid cloud + AI. IBM today has a market cap of $207.5 billion. It was once the most valuable company in the world, then nearly died in every technological revolution, and every time climbed back out of its coffin.
Let me tell their story first. After that, we'll talk about why "surviving every extinction" is harder than "leading a single era."

Watson Sr. joined the Computing-Tabulating-Recording Company (CTR) in 1914 — it had 1,300 employees at the time. Watson Sr. was a sales genius with a messiah complex. He forbade employees from drinking — alcohol-fueled decisions had wrecked his own previous job. He required all male salesmen to wear dark suits and white shirts — "what you're selling isn't a machine, it's a faith in precision."
On his wall hung only one word — THINK. That single word has been printed on the walls of every IBM office, from Manhattan to Tokyo, for over 100 years.
During the Great Depression of the 1930s — IBM did not lay anyone off. Watson Sr. made a decision that horrified the board: he shifted the factory's production lines from tabulating machines to stockpiling parts — hoarding. He bet that Roosevelt's New Deal would generate a massive ocean of government data-processing demand — Social Security, labor statistics, the census. He bet right. After the Social Security Act passed in 1935, IBM became the sole designated supplier of data processing for the U.S. government.
Then he made a nearly fatal error — he underestimated the electronic computer.

In the late 1940s, IBM was still selling mechanical punch-card tabulating machines. A company called Remington Rand sold the first commercial electronic computer — the UNIVAC I — in 1951. Every newspaper headline read "IBM is the next dinosaur headed for extinction."
Watson Sr.'s health began to decline, and he handed the company to his son — Thomas Watson Jr.

When Watson Jr. took over his father's company, he made an even wilder bet than his father's. He poured 60% of IBM's annual revenue into a project called System/360 — with a total investment exceeding $5 billion (the equivalent of over $40 billion today). If it failed — IBM would simply vanish.
System/360 was a "computer family" that unified all computers, large and small, under a single architecture — you bought a small model, upgraded to a larger one, and didn't need to rewrite all your software. This was the first time anyone in the computer industry had created a "compatibility" system.
On the day of its launch in 1964, Watson Jr. stood on stage with his hands shaking — he wrote about it later in his memoirs. He chose bright red cabinets, a color IBM had never used before, and placed them before the assembled media. Fortune magazine later called it "the greatest personal bet in business history." System/360 became the core chassis of the entire mainframe era. By the 1970s, 70% of all computers in the world were IBM mainframes. And one of those red cabinets now hangs in the Smithsonian.
Watson Jr. later told a reporter: "My father taught me how to sell. But I had to tear down with my own hands the product lines my father had spent his entire life building. Be your own gravedigger."
Watson Jr. retired in 1971. IBM entered a comfortable period where it felt it "could never make a mistake" — and then, in the 1980s, the personal computer revolution arrived. IBM invented the first IBM PC — but outsourced the operating system to Microsoft and the chip to Intel. This was the most famous "suicidal outsourcing" in business history — because it turned IBM into a replaceable assembler.

In the 1990s, IBM nearly disintegrated. In 1993, the company lost $8.1 billion — the largest annual corporate loss in U.S. history at the time. The board poached Lou Gerstner from RJR Nabisco — he sold cookies and cigarettes and knew nothing about computers. Everyone said he couldn't save IBM.
Gerstner did one thing: he rejected every investment bank's proposal to "break up IBM." He said: "IBM's value doesn't reside in any single business unit — it resides in being the only integrator that can bind together every IT layer — from chips to software to consulting. Breaking it up would be like killing an elephant and expecting its scattered pieces to run on their own."
He transformed IBM from "selling mainframes" to "selling IT solutions and global services" — pivoting the entire company from hardware manufacturing to a services company. And IBM came back to life, walking into the 21st century.
In 2018, IBM bought Red Hat for $34 billion — betting on hybrid cloud. Then it bet on quantum computing and the next-generation enterprise platform of the Watson AI era.
It is no longer the largest computer company on Earth. But it is still at the door of every new era — in line, waiting to enter.

Three things.
First thing: THINK is not a slogan — it's a decision-making culture.
Watson Sr. forced all disputes to be resolved within a single decision framework — "Are we thinking about what the customer needs 10 years from now, or the quarterly report 10 months from now?" He required all executives, when objecting to someone, to say the word "Think" — reminding themselves they were not venting emotion, they were using their heads.
Does your team have such a "trigger word"?
Second thing: Bet the company on one product — and simultaneously ensure that product is compatible.
System/360 was a home-run bet — but the bet wasn't on one model, it was on an architecture compatible across all models. Even if the large model didn't sell, the small ones could still grow the software ecosystem.
Is your "big bet" — designed so that "if one part loses, the seeds can still grow another tree"?
Third thing: Giants don't die from a single blow. Giants die from refusing to let go.
Watson Sr. refused to abandon mechanical tabulating machines, and nearly missed electronic computers. Then Watson Jr. won with the 360 bet — but IBM later refused to let go of the mainframe and nearly missed the PC. Then the PC tripped it up too. Then it caught its breath again in services. And did it all over again in the cloud computing era.
Every time it "nearly died" — it wasn't for lack of technology. It was because it refused to let go of its last success.

IBM is 114 years old this year. It hasn't died. It competes with cloud-native companies 80 years younger than itself and technologies a full century younger. Jensen Huang and Satya Nadella are both its competitors — and also, on certain projects, its partners. That a 114-year-old company can still sit at this table — that in itself is one of the most astonishing facts in all of business.

← Previous Contents Next →
Chapter 51 PepsiCo

1893, New Bern, North Carolina. A pharmacist named Caleb Bradham mixed up a jar of brown sugar water in his basement. He called it "Brad's Drink" — for indigestion. His neighbors said the burp after drinking it felt great. He renamed it Pepsi-Cola. Today PepsiCo has a market cap of $207.4 billion.

Not an invention patent. Not a technological disruption. Not a first-mover advantage.
Just Pepsi. In nearly every decade, it struggled for survival in the shadow of Coca-Cola. It climbed back from bankruptcy. In the 1930s, it offered to sell itself to Coca-Cola — Coca-Cola didn't want it. Then a candy store owner and a marketing genius picked it up and, facing the longest red shadow in the world — fought a hundred-year war.
Pepsi has never outright dominated Coke in a blind taste test — but it did something far smarter: it refused to fight on the battlefield of "whose recipe is better." It moved the war to "who is younger, cooler, and better at marketing."
This story is about — when every rational battlefield has been occupied by the king, where the number two should go to win.

When Caleb Bradham invented Pepsi, Coca-Cola had already been alive for seven years. For the next 100 years, Pepsi had only one label — the chaser.
In 1931, Pepsi went bankrupt. The Great Depression. Bradham lost everything. A candy store chain owner named Charles Guth took over the mess. Guth did the first thing that truly made Coca-Cola uncomfortable — he put Pepsi into 12-ounce bottles, the same size as Coke, but sold it for just 5 cents — the exact same price as Coca-Cola's 6-ounce bottle.
This was the "more volume, same price" assault — consumers held up two glass bottles of the same size, 6 ounces on the left, 12 ounces on the right. In the depths of the Great Depression, the cheaper option won.
In the 1950s, Pepsi was no longer the poor man's cola — it needed a complete image revolution. The "Pepsi Generation" marketing campaign was the first in history where a consumer brand said: "I'm not selling taste — I'm selling your sense of which kind of person you belong to."
In 1964, it built a model that every beverage brand would later imitate — signing music megastars (Michael Jackson, and later Beyoncé, Cardi B). The endorsement fees were outrageously high — but Pepsi figured out the key insight: the profit margin on a single bottle of cola is just too small. You have to push foot traffic to the absolute limit — and a superstar can, in a three-minute song, make millions of teenagers feel they are "Pepsi people."
Then in 1975, Pepsi pulled off one of the most famous marketing ambushes in business history — the Pepsi Challenge. Two unlabeled white cups placed at the entrance of a supermarket — consumers drank blindfolded and pointed to whichever cup tasted better. Cameras filmed the whole thing. Ordinary consumer after ordinary consumer, earnestly on camera — "This one." Then the cover lifted: Pepsi.
Coca-Cola was thrown into panic by this street-corner blind taste test — in 1985, they launched the New Coke formula, and it triggered a catastrophic reaction. That year, Pepsi briefly surpassed Coca-Cola in U.S. domestic sales. The New Coke debacle was the greatest brand defeat in Coca-Cola history — and the instigator was a blind taste test with white supermarket cups.

Then Pepsi did the most "traitorous" thing any beverage company could do — it stopped being just a soda company.
In 1965, Pepsi-Cola merged with Frito-Lay — the parent company of Lay's, Doritos, and Cheetos. The logic of the merger was this — snacks and beverages are always purchased in the same retail context, but their supply chains and raw material sourcing have almost no overlap. So Pepsi became a "snacks + beverages" dual-engine conglomerate.
Today, roughly 60% of PepsiCo's profits come from snacks — Pepsi-Cola itself doesn't even generate as much revenue as Lay's potato chips.
When global soda consumption began declining from the mid-2000s onward — Pepsi didn't have to scramble entirely on Zero, sparkling water, and sports drinks to survive, the way Coca-Cola did — it had an entire snack empire propping up its stock price. Coca-Cola, meanwhile, has always been a pure beverage company, with no snack hedge.
This is the most profound strategic decision in Pepsi's history — don't be a pure beverage company, be the "optimal dual-category portfolio." In today's global anti-sugar wave and the panic over GLP-1 weight-loss drugs, this hedge is worth more than ever.

Three things.
First thing: If you can't surpass your rival on "better" — switch the dimension of competition.
Pepsi has never decisively won on "whose cola is better." But it switched dimensions — "who understands young people better," "who owns a broader snack corridor." On which dimension is your current competitor crushing you? Can you switch to a battlefield they're bad at?
Second thing: The blind taste test is the weapon number one fears most.
The Pepsi Challenge pulled Coca-Cola down from the brand altar to the bottom of a white cup — inside that cup, a hundred years of red faith vanished. Does your industry have the opportunity for a "naked-cup blind test" — strip off the label, does your product actually hold up?
Third thing: Dual-category hedging.
Pepsi hedged the systemic risk of "beverage decline" — with "snack growth." When humans eat salty chips — they crave something sweet or carbonated. When humans drink sugar-free beverages — their other hand is grabbing a fistful of Doritos. This is a designed compensatory loop.
Does your product portfolio — have a natural hedge? When one category is abandoned by the era — does another automatically rise to take its place?

That little pharmacy in New Bern, North Carolina, where Pepsi was born — is now a retail outlet for a Pepsi historical museum. Visitors can buy a bottle of "Brad's Drink" replica — made with cane sugar instead of high-fructose corn syrup. Inside that bottle isn't the war between Pepsi and Coca-Cola — it's the story of a second place no one ever believed could win, that found the next place to turn the tables every single time it was about to die.

← Previous Contents Next →
Chapter 52 Blackstone

1985, two Lehman Brothers partners each put up $200,000 and rented a small office in Manhattan. At the first morning meeting, one looked at the other and said: We will never do a hostile takeover. There are already enough people in this world stealing from each other. What we're going to do is — buy companies, help them become better, and then sell them. Today, Blackstone manages over $1 trillion in assets and has a market cap of $202 billion.

Not high leverage. Not hostile takeovers. Not breaking up companies. Not layoffs and restructurings.
Just Steve Schwarzman and Pete Peterson. Two bankers who left Lehman Brothers. When they founded Blackstone, they had no fund, no provable track record. Schwarzman personally took trains to knock on the doors of every pension fund and university endowment. He was rejected over 400 times.
They insisted on steering clear of hostile takeovers. And then they turned "friendly buyouts + operational improvement" into the largest alternative asset management machine in the world.
Let me tell this story first. After that, we'll talk about why "being a decent person" can also become an invincible moat — even in the world of finance.

Steve Schwarzman. Born in Philadelphia, his father ran a curtain shop. Yale undergrad, Harvard MBA. He rose to head of the M&A department at Lehman Brothers — it was there that he formed his view on the core value of acquisitions: M&A shouldn't be a war. M&A is a capital engine that "improves the efficiency of a machine, then sells it when it's running better." Pete Peterson was the former CEO of Lehman Brothers, 20 years older than Schwarzman, the opposite in temperament — Peterson was gentle and low-key, Schwarzman was ferociously impatient. But on one principle, the two had an ironclad consensus: no hostility. It came from the chaotic battle when Lehman Brothers was sold to American Express in 1984 — the animosity and personal destruction it stirred up across all of Wall Street horrified both of them. They swore — every deal at the new firm must be one the target company's board agreed to, welcomed, and believed could be a win-win.
In 1985, no one on Wall Street believed in this approach. The trend at the time was the "corporate raider" — using junk bonds to force acquisitions, break companies apart, lay people off, and take a big bite of profit. Schwarzman and Peterson insisted on friendly buyouts. Bankers laughed at them: "These two want to be paladins."
For the first 18 months, Blackstone didn't close a single M&A deal. Schwarzman took trains every day from New York to wherever a pension fund manager would grant him five minutes — after 400 rejections, Prudential Insurance finally gave the first investment commitment.
Then Blackstone grew, slowly. It used friendly buyouts to acquire hotels, office properties, chemical distributors, and music catalog rights. Each time, the CEO of a portfolio company would say on referral calls — "They didn't come to take my stuff. They came to help me figure out which pieces could be sold for more." This became Blackstone's unwritten advertisement.

In June 2007, Blackstone went public — at the time, the largest private equity IPO in U.S. history. One month after listing, the first shockwaves of the subprime crisis hit. Blackstone's stock price fell from $31 on its debut day all the way down to $3 in early 2009. Over those two years, at every investor call he attended, Schwarzman repeated again and again: "We did not invest in subprime. Our assets have no direct exposure to subprime mortgages. The average duration of our liabilities is longer than the life of our funds. We will survive this cycle."
After 2009, Blackstone not only survived — it used the crisis to acquire vast quantities of panic-sold commercial real estate and distressed debt at extremely low prices. From 2012 to 2019, it grew into the largest commercial real estate owner in the world. During the brief panic of COVID in 2020 — it once again used its "long-duration fund" advantage to catch forced-sale asset packages. By 2025, its assets under management exceeded $1 trillion — it's not the biggest on Wall Street — but its "friendly + long-term capital" structure turns every economic recession into its buying season.

Three things.
First thing: Reject hostile takeovers — and attract more deals instead.
When founders know that Blackstone won't kick them out after selling the company — they proactively call and ask, "Would you like to look at the strategic M&A cases being discussed in our boardroom?" This became a deal-flow engine that people willingly came knocking for. Compared to reading in the newspaper that your company is about to be forcibly seized — selling to a "friendly hand" is a relief.
When you attract clients — is "you won't lose your job / relationship" your greatest invisible selling point?
Second thing: The first fund is the hardest — 400 rejections before the first yes.
Schwarzman took trains and knocked on 400 doors before getting Prudential's first commitment. Everyone's first fund is like this — you're not buying assets, you're selling your own credibility.
Do you have a "400-door record" — and if you haven't yet knocked on 400 doors, have you actually even started selling in earnest?
Third thing: Permanent capital + long duration = counter-cyclical harvesting machine.
Blackstone's core structure has never been short-term hedges — it's 10-year funds. Holders cannot redeem at any time. This means during market panics, it faces no redemption pressure — instead, it is the only one still making bids.
Can your current capital source allow you to become that "only person in the room still raising a paddle" during a storm?

Schwarzman wrote his autobiography, What It Takes. On the first page, he wrote about his father in that Philadelphia curtain shop, sorting and shelving curtain fabric every Saturday morning — defining one thing over a lifetime: decency.
Blackstone's logo is a shield, black lettering on a white background. Schwarzman gave the design to the designer verbally — "Because we are reliable partners who can afford to pay compensation."

← Previous Contents Next →
Chapter 53 Verizon

2000, Bell Atlantic acquired GTE for $64.7 billion — then did something that left Wall Street analysts utterly baffled. They threw both names in the trash and coined a new word: Verizon — a fusion of "Veritas" (Latin for "truth") and "Horizon." They said this new network was not a phone company — it was the "information horizon." Today Verizon has a market cap of $200 billion.

Not a technological first. Not a content ecosystem. Not a low-price strategy. Not international expansion.
Just "network quality above all else." From full LTE coverage in the 2000s to 5G Ultra Wideband in the 2010s, Verizon has consistently held the number one or number two position in U.S. mobile network quality rankings. Its strategy has gone from transmitting voice in the Bell era — to transmitting hundreds of megabits per second of data today — the medium changed, but that belief in "signal reliability" never did.
When Verizon first laid fiber to the home, its peers mocked it as "capital suicide" — the coaxial cable companies said "coax is good enough for the next decade." Then everyone followed it and started digging fiber trenches. Verizon didn't predict the future — it paved the future with fiber-optic cables in the ground.

Three things:
First thing: How much organizational courage does it take to throw away the names of two century-old legacy brands — and replace them with an artistic fusion of "truth + horizon"?
Second thing: The quality moat of heavy assets — when every fiber-optic strand is one you dug the trench for and buried yourself, competitors can only lease — and you can technically stay one step ahead of the lessees in iteration, forever.
Third thing: When everyone in the industry says "the existing infrastructure is good enough" — be the one who lays fiber first.

← Previous Contents Next →
Chapter 54 Analog-Devices

1965, two MIT graduates — Ray Stata and Matthew Lorber — founded a company in an old garage in Cambridge. They didn't make digital chips, or DRAM, or microprocessors. They made something considered "old-fashioned" in the all-digital age — analog chips. Turning the real world's temperature, pressure, and sound waves into precise digital signals. Today Analog Devices (ADI) has a market cap of $199 billion. Without ADI's analog-to-digital converters, your phone's microphone couldn't hear your voice, your car's radar couldn't sense the distance to the car ahead, and your MRI scan image would be a blur of white noise.

Not the digital revolution. Not Moore's Law. Not CPUs. Not AI chips.
Just "the bridge connecting the physical world and the digital world" — the analog-to-digital converter (ADC). Ray Stata was the engineer who, while everyone else charged toward digital, walked the other way — toward analog. Because he saw the most fundamental physical limit of the digital world: the real world is a continuous curve — and chips only understand 0 and 1. Someone had to build the translator. ADI made that "translation" the most precise on Earth.

Three things:
First thing: Always be the "translator" — not the lead actor, but the choke point that every lead actor depends on.
Second thing: Analog chip design is a craft, harder than digital — senior engineers' experience values cannot be auto-generated by EDA tools.
Third thing: As sensors enter every car and every factory — ADC usage grows exponentially, and ADI controls the highest-precision tier.

← Previous Contents Next →
Chapter 55 McDonalds

1954, a 52-year-old milkshake machine salesman drove three days from Chicago to San Bernardino, California. He'd received a strange order — a hamburger stand had bought eight milkshake machines. Eight machines meant they were making 40 milkshakes at once. He had to see it with his own eyes. At noon that day when he arrived, he saw something he'd never seen in his life — a hamburger going from order to handout, compressed from 15 minutes down to 30 seconds.

Not inventing the hamburger. Not inventing fast food. Not inventing the franchise model.
Just Ray Kroc. 52 years old. Sold paper cups, sold mixers — sold things his whole life. He saw the McDonald brothers' work counter — and then did something the brothers had always wanted to do but never could: replicate that 30-second kitchen at every crossroads across the world. And then he "took" the company away from the founding brothers. Today McDonald's has a market cap of $197 billion.
Let me tell his story first. After that, we'll talk about why the "founder" and the "scaler" are often two completely different people — and the latter frequently walks away with all the money.

In 1940, brothers Richard McDonald and Maurice McDonald opened a drive-in restaurant in San Bernardino. Business was good. But three problems tormented them: first, the drive-in model attracted juvenile delinquents and chaos; second, food took too long (over 15 minutes), and waiting customers got irritable; third, labor costs were extremely high — short-term waitstaff had high turnover, trained and ran.
In 1948, they made a decision that looked like suicide in the restaurant industry at the time — they shut down the restaurant. When it reopened, the doors were sealed shut — no more letting customers wait in their cars, no more carhops delivering food curbside. Customers had to get out of their cars, line up at the window, and order themselves.
Then they turned the kitchen into a "hamburger assembly line" — every person performed only a fixed, repetitive motion. The burger flipper only flipped patties, the sauce squeezer only squeezed sauce, the wrapper only wrapped. And all hamburger toppings were fixed — no customization. There was no "made your way" back then. This was total standardization — speed raised to 30 seconds. Price dropped from 35 cents to 15 cents.
The McDonald brothers were process geniuses. But they were not expansion geniuses.
Ray Kroc was.

Ray Kroc was 52 that year. A salesman who started as a World War I ambulance driver. He'd sold paper cups, then moved on to milkshake machines — the Multimixer — a machine that could blend five milkshakes at once. He'd crisscrossed America for over a decade, selling them to restaurants. When he saw the McDonald brothers running eight Multimixers simultaneously — he stood at the edge of the parking lot, watching the flow of the line, and his mind was already spinning at high speed: "If this store were in Chicago, on the highways, on the East Coast..."
He walked up and asked the brothers: Why don't you open more?
The brothers said: We don't want to. Too much hassle. And no one can do it the way we do — the first franchise store, just an hour away, already couldn't maintain quality control.
Kroc tossed and turned in his motel room that night. The next day he knocked on the door. He made a proposal — sell him the franchise rights to McDonald's. The brothers would keep running the original store. He would fly all over America selling licenses — using standardized training, a unified menu, unified procurement, and undercover quality inspectors making spot checks at every store to guarantee quality.
This was a "central legal + shared procurement + unified brand" licensing model — later copied by every fast-food chain. But at the time, no one had ever done it at anything close to this scale.
In 1961, the brothers felt they'd made enough. They sold the brand name, the logo, and all trademark rights to Kroc for $2.7 million. That wasn't a small sum — but compared to what McDonald's would later be worth, it may be the lowest sale price for a "future patent right" in business history.

After Kroc gained full control, he did two things the entire industry mocked.
First: in 1961, in the basement of a McDonald's in Elk Grove, Illinois, he founded Hamburger University, turning French fry making and burger flipping into a curriculum with a certificate. He personally designed the McDonald's operations manual — over 75 pages thick. The most important thing in chain restaurants isn't great chefs — it's an operating guide that anyone can open and replicate across 12 shifts in a single day.
Second: he stopped making money from licensing fees. The traditional chain licensing model was selling regional licenses + collecting royalties. Kroc's real profit didn't come from a cut of franchisees' revenue — it came from buying the land under each store site, then subleasing the storefront to the franchisee. McDonald's Corporation was a "real estate company"; the product was hamburgers.
This was Kroc's most underrated business innovation — digging up land, buying land. He bound every franchisee to a McDonald's real estate lease — if they didn't follow the standards, the lease would be voided immediately, and they wouldn't even own the bricks in the walls. This turned quality control into "terms of a lease" rather than "friendly suggestions."

Stop here for a moment.
Are you a "McDonald brother" or a "Ray Kroc"? The person who invented the hamburger opened a few stores in his lifetime. The person who sold the brand put a hamburger onto plates in 120 countries. Most founders are the former. An extremely rare few are the latter. You don't have to be both — but you must be honest with yourself: are you inventing something new — or scaling an existing good idea?
Ray Kroc was not the inventor of the hamburger. But he turned that 30-second kitchen into a shared eating habit for all of humankind.

Three things.
First thing: Process standardization = a recipe that any human anywhere in the universe can replicate.
Did the McDonald brothers make a better burger? No. What they did was break the act of "making a hamburger" into irreducible pieces, then filmed it as a sequence of steps anyone could repeat.
Can your core skill — be broken down to the point where "anyone who opens the manual can get it in three days"?
Second thing: The money isn't made from the hamburger — it's made from the land.
Kroc turned McDonald's into one of the largest holders of chain-restaurant real estate in the world. In the prospectus, he said plainly: "Our primary business is real estate. The secondary business is hamburgers."
Do you have a revenue structure where "what looks like the side business is actually the main business"?
Third thing: 52 is not too late.
When Ray Kroc encountered the McDonald brothers, he was already past the age when most people tell themselves "this is how my life is going to be." He was driving his Buick thousands of miles with diabetes, arthritis, and a body that had nearly died in surgery.
Do you feel "it's already too late" — are you older than the 52-year-old Kroc?

Ray Kroc left a quote on the wall of McDonald's headquarters: "Nothing in the world can take the place of persistence. Talent will not — nothing is more common than unsuccessful men with talent. Genius will not — unrewarded genius is almost a proverb. Education will not — the world is full of educated derelicts. Persistence and determination alone are omnipotent. — Press On."
The year he died, McDonald's had just passed 3,000 stores worldwide. Today, over 40,000. Every single one fries its French fries the exact same way.

← Previous Contents Next →
Chapter 56 NextEra-Energy---新纪元能源

1925, Florida was still mostly swampland and orange groves. A small power company called FPL (Florida Power & Light) was founded in Miami — its earliest generating equipment was a diesel generator salvaged from a sunken ship and retrofitted. A hundred years later, its parent company NextEra Energy has a market cap of $196.9 billion and is one of the world's largest solar and wind power producers — at a time when nearly every oil company is still "discussing" a low-carbon transition, this century-old Florida utility has already built renewable energy into the cheapest source of electricity on Earth.

Not an energy revolution. Not policy subsidies. Not environmental idealism.
Just Jim Robo. After taking over the company in 2006, he made a decision that terrified every peer in the power industry — he bet virtually all of NextEra's new generation investment on wind and solar, while pulling out of coal at scale. This wasn't because he had a moral obsession with climate change — it was because he ran the levelized cost-of-energy curves on a spreadsheet over a 20-year horizon, and saw that the cost of wind and sun was falling irreversibly.
When natural gas prices spiked in the 2020s, NextEra's wind farms in Texas were still putting out power at a cost of 2 cents per kilowatt-hour. Its subsidiary FPL in Florida still maintains some of the lowest residential electricity rates in the nation — while Florida summers can bake pavement to 70 degrees Celsius and air conditioners run 24 hours a day.
Robo's logic comes down to one thing: the price of fossil fuels jumps up and down — wind and sun are free. Once the wind turbine is built, the fuel cost is a constant zero. And a "power plant with zero marginal cost" is the ultimate moat for any utility.

Three things:
First thing: The transformation wasn't because it was "right" — it was because the math showed the cost would be zero after 20 years.
Second thing: Wind and solar are free hedges against fossil fuel — not an environmental badge, but the lowest-volatility assets in the realm of capital discipline.
Third thing: A 100-year-old utility can become a new-energy giant — provided the CEO is willing to voluntarily scrap old assets while they're still making money.

← Previous Contents Next →
Chapter 57 Disney

1928, he lost his first successful cartoon character — Oswald the Lucky Rabbit. It wasn't the market that defeated him. It was his distributor, who hid a clause in the contract: the copyright to Oswald — did not belong to him. He sat on the train from New York back to Los Angeles and drew a mouse on a napkin. Today that mouse is worth $182 billion.

Not creative genius. Not technological breakthrough. Not financial engineering. Not luck.
Just Walt Disney. A farm boy from Kansas City. He wasn't the most talented animator in Hollywood. But he understood one thing more clearly than anyone else: cartoons aren't drawn for children — they're drawn so the whole family can laugh together in the same dark room.
Then, with a Mickey Mouse, a Snow White, and a "theme park" the likes of which had never existed on Earth — he turned "entertainment" from a product into a world you could step inside.
Let me tell his story first. After that, we'll talk about why "turning a single transaction into a lifetime of memory repurchases" is Disney's deepest business moat.

Walt Disney was born in Chicago in 1901. His father was a harsh manual laborer and religious fanatic — the kids in their family were roused at 3:30 every morning to deliver newspapers, after which Walt would go to school (and often fall asleep in class). In his youth, the only place he could escape reality was on drawing paper — on any scrap he could find, he drew farm animals, a singing sun, fairy tales. During World War I, he lied about his age to drive ambulances in France — and covered the ambulance from end to end with cartoons.
In 1923, he arrived in Hollywood with $40. He and his brother Roy started an animation company in their uncle's garage. Their first series was called the Alice Comedies — a live-action little girl appearing alongside cartoon animals. The films sold purely because, in that era, "nothing else looked like this."

In 1927, Walt created a rabbit — Oswald. It was the first cartoon character that truly gave the Disney studio stable income. The copyright was registered under the distributor, Universal Pictures, and the contract contained a phrase Walt couldn't parse — "work-for-hire." In 1928, when Walt asked for a raise, Universal kept the copyright — and poached his core animation team.
This was the greatest betrayal and "technical robbery" of Walt's life. In a train car heading back to Los Angeles, he wiped away tears, then pulled a pen and a napkin from his pocket. He drew a round-eared mouse — originally named Mortimer. His wife Lillian said: "That name is too stupid. Call him Mickey."
And so — Mickey Mouse was born.
But it was the third thing that transformed Walt from "animation studio boss" into the eternal Walt Disney.

In November 1928, he released Steamboat Willie — the world's first animated short with fully synchronized sound and picture. At the time, all films were silent, and cartoons were a five-minute filler — when Walt made the mouse whistle in the screening room, perfectly matched to the various sounds on the boat, every theater manager present rose to their feet and applauded. The sound wasn't laid over the picture — the picture, the music, and the sound effects had been a unified composition from the very first stroke of the pen.
This wasn't "adding sound effects." This was reinventing the relationship between animation and the human brain.
Then he made a bet that all of Hollywood laughed at — he wanted to put an 83-minute animated feature into movie theaters. At the time, no one was willing to sit in a theater and watch 83 minutes of cartoons — people watched cartoons while waiting for the main feature. Snow White premiered in 1937. When it ended, the entire audience — including Hollywood royalty like Cary Grant and Shirley Temple — stood up in tears and applauded. A cartoon had, for the first time, moved people to tears. And it grossed $8 million at the box office — in 1937, that was a fortune.
Walt used the Snow White money to build the Burbank Disney studio — and then, in 1955, made the bet that was mocked most of all by engineers and bankers: build a theme park. Every banker said: "Who's going to drive their family for hours into the middle of a cornfield to sit in a fake elephant boat?" Walt bought up an orange grove in Anaheim, Orange County, California.
It was the first Disneyland on Earth — not an amusement park with carousels and Ferris wheels. It was a "storybook you could walk into" — inside the park, every trash can matched the theme of its zone. The windows on Main Street were polished to reflect the morning sun. His "Imagineers" — Dream Engineers — crossbred film set techniques with mechanical engineering to create a world where you forgot you were standing in line for a ride.
Then Walt Disney World rose in Florida — then Tokyo, Paris, Hong Kong, Shanghai.

Walt died of lung cancer in 1966, at age 65. Disney World was still under construction. Decades after his death, Disney bought George Lucas's Lucasfilm for $4 billion, bought most of 21st Century Fox's assets for $71.3 billion — and put Marvel, Star Wars, and Avatar onto Disney+ streaming. Every generational shift, someone declares "Disney is dead," and every time, its IP matrix weaves together a new childhood for a new generation.

Stop here for a moment.
What you sell — is it "used once and it's over" — or can someone "walk in and stay a lifetime"? Walt took a three-minute cartoon short, made it into an 83-minute story, then made it into a theme park you could walk into, then made it into a streaming service you renew every year.
He wasn't selling animation. He was continuously thickening the "density of the narrative medium" of a single worldview. Can you find a way — to turn your "thin, single-use product" into a realm people can enter again and again?

Three things.
First thing: Never define animation by the limitations of "cartoons."
Before Steamboat Willie, animation was silent vaudeville. After Walt, animation had symphonies, operatic plots, and deaths that made people cry. He wasn't improving a category — he was letting it grow up.
Is your product — still being treated like a child by the unwritten "rules" of your whole industry — and are you, in fact, the only one who can give that category its full adult stature?
Second thing: Make a business for the whole family — not just for kids.
Disneyland is not a children's playground. It's "the place where adults are permitted to become children again." When your product can immerse a six-year-old and a sixty-year-old at the same time — your addressable market is the entire human lifespan.
Does your product have a feature that lets the "inner child" of an adult run free, safely?
Third thing: Losing the first one isn't the end — it's your first lesson in intellectual property.
After losing Oswald the Rabbit, Walt wrote into every contract, on the second line from the top: "Character copyright belongs to Walt Disney Studios." He then built what may be the entertainment industry's most rigorous intellectual property protection system.
Have you — "because you got robbed once — built a system that ensures no one can ever rob you again"?

Mickey Mouse is 98 years old this year. His big ears are still round, still black-and-white shorts, still those big yellow shoes. Children born all over the world don't yet know their own names — but they can recognize those two round ears.

← Previous Contents Next →
Chapter 58 Boeing

1916, Seattle. A Yale-educated timber merchant sat inside the first seaplane he'd built himself — a biplane stitched together from spruce wood, linen, and piano wire, with a wooden propeller and an engine converted from a boat motor. He test-flew it himself. The moment it touched down on the surface of Lake Union, he said to his partner: We're going to build the best aircraft in the world. Today, Boeing has a market cap of $182.2 billion, and at any given second, thousands of Boeing jets are flying through the skies across the globe.

Not an aeronautical engineer. Not a fighter pilot. Not an MIT lab. Not a government contract.
Just Bill Boeing. A wealthy timber merchant. In 1910, he watched an air show in Los Angeles for the first time — the biplane wobbled around in the sky, circling three times. Most people saw "flying is thrilling." Bill Boeing saw a pile of uncalibrated, unstandardized, haphazard parts — like something out of a toy store — somehow getting airborne. He said: "We can do this a whole lot better."
Then, with timber, a boat engine, and a friend who did naval engineering, he built the first B&W seaplane in a boathouse. Today every Boeing 787 Dreamliner's fuselage is woven from carbon fiber — but its first plane was planed out of spruce boards from his own timber stands.
Let me tell his story first. After that, we'll talk about what it really takes to turn an immature toy into the most reliable industrial product in the world.

William Edward Boeing. Born in 1881 in Detroit, from a wealthy family; his father was a mining magnate. After graduating from Yale, he moved to Washington State and went into the timber business — bought up vast tracts of forestland, sold spruce to shipyards. He was the kind of man who wore a three-piece suit on the sawmill floor, inspecting spruce grain patterns through the wood dust — his obsession with raw materials came from here.
After watching the air show and taking a test flight in 1910, his assessment of the airplanes of the time was "unsafe box kites" — pieced together in barns by individual craftsmen with no engineering standards, no two planes exactly alike. In 1915, he started building his own plane in a boathouse on Seattle's Lake Union, using hand-selected straight-grain spruce from his own timber stands — not for furniture, but for cutting wing ribs. His partner Conrad Westervelt was a naval engineer, in charge of converting boat engines into aircraft engines. The first B&W plane — green lacquer, wooden fuselage, water takeoff and landing — was sold to a New Zealand flying school.
Then, in 1917, America entered World War I. Boeing landed the Navy's first order for training planes — and built a real aircraft factory from scratch.

But what truly made Boeing wasn't World War I — it was the bet on airmail during the Great Depression of the 1930s, and the B-17 and B-29 bombers in World War II. Boeing conquered three core technologies simultaneously — long range, pressurized cabins, and all-metal stressed-skin construction — and then, after the war, transferred the B-29's wing and pressurization technologies into commercial aircraft.
In 1954, Boeing made a truly insane bet — with zero customer orders, it poured all its profits into building a prototype jet airliner — the Boeing 367-80 (the "Dash 80"). The cost exceeded $16 million. The entire company's annual profit was only a few million. If this bet failed, Boeing would be forced into a total sale or liquidation.
In 1955, during Seattle's Gold Cup hydroplane race, the Dash 80 performed a low-altitude barrel-roll turn that made every aviation executive in attendance spill their drinks. Then Pan Am placed an order for 20 Boeing 707s — the jet airliner age had begun. The 707 is the aerodynamic prototype for every swept-wing jet airliner since, including every Boeing and Airbus model flying today.
And then the 747 — born out of a conversation with Pan Am founder Juan Trippe: "What if I want to fly 400 people across the Pacific?" Boeing's chief engineer Joe Sutter led the "Incredibles" team from a blank sheet of paper to first flight in under four years — the 747's "hump" upper deck was the product of a compromise, and it ended up becoming the most recognizable aircraft silhouette in the world.

But here is a dark chapter that must be written.
After Boeing merged with McDonnell Douglas in 1997, its corporate culture shifted from "if the engineer says it can't fly, it doesn't leave the factory" to one driven by shareholder returns and cash-cycle cadence. In 2018 and 2019, two 737 MAX aircraft crashed, killing 346 people. The investigation found that a software system called MCAS was designed to rely on a single sensor input — and critical information about the system was not adequately disclosed to pilots and regulators. Engineers had boasted in internal emails about fooling the FAA's oversight "at the lowest cost." The MAX was grounded globally for over 18 months — one of the worst safety crises in aviation history.
Boeing went from an engineering faith built on "inspecting every wood grain amid the sawdust" — to having its own internal emails become evidence in a criminal investigation. Then COVID, the 787 delivery freeze, and the failed crewed maiden flight of the Starliner spacecraft — have left it still struggling to this day.
At the same time, roughly half of every commercial airliner landing at any airport on the planet is a Boeing, and the other half is an Airbus. Boeing's global fleet supports the entire long-distance movement of the human race.
This story tells everyone at once: the cost of losing engineering faith — is not fines. It's trust. And once trust is gone, re-testing it is ten thousand times harder than designing a new aircraft model.

Three things.
First thing: When there are no customer orders — build the extreme prototype first.
The Dash 80 was one of the most dangerous self-funded R&D projects in business history. But that prototype became the physical ancestor of every Boeing airliner for the next 50 years. Are you willing, with no orders in hand — to first build a prototype that is ten times beyond existing standards?
Second thing: The skeleton of trust isn't a compliance form — it's the cultural habit of "if the engineer says it can't fly, it can't fly."
When Boeing's chief engineer vetoed an aircraft, the VP of sales had no authority to object. Then one day, the voice of the sales department drowned out engineering — and MCAS was slipped outside the manual. Right now, are you letting the voice of the "safety / quality / principles" department be drowned out by the "revenue" department?
Third thing: Look at metal through the eyes of wood.
Boeing's founder judged materials through the rings of a tree. Even when he later used all-metal and carbon fiber — he forever asked suppliers: "Show me the uniformity data for this batch." In your physical raw materials — whether code, steel, or content — does anyone run their hands along every grain the way Boeing looked at spruce?

Bill Boeing passed away in 1956 — after the 707's first flight, but he lived to see the beginning of the jet age. In his will, he gave away the vast majority of his fortune. His first boathouse on Lake Union — now next to Seattle's Museum of Flight — still has waterbirds landing on the lake's surface. They don't know that beneath the water lie the wood shavings of the first B&W seaplane.

← Previous Contents Next →
Chapter 59 Amgen---安进

1980, a venture capitalist named Bill Bowes and several scientists founded AMGen (Applied Molecular Genetics) in Thousand Oaks, Southern California. They had no drug molecule of their own, no factory — only a theory: recombinant DNA technology could "print" human protein drugs from E. coli bacteria. The entire industry said it was science fiction. In 1989, Amgen launched Epogen — recombinant human erythropoietin, enabling kidney failure patients' bodies to grow their own red blood cells — from that point on, external blood transfusions were no longer the only option. Today Amgen has a market cap of $181.3 billion.

Not pharmaceutical chemistry. Not herbal extraction. Not immunosuppression. Not traditional drugmaking.
It's "take a protein your body already has but doesn't have enough of — ferment it out using genetically engineered bacteria, then inject it back into you." This is biotech pharmaceuticals — not chemical drugs. Amgen's first-generation scientist Fu-Kuen Lin spent years fishing the EPO gene out of the human genome — then spliced it into Chinese hamster ovary cells — turning the cells into micro-factories for protein drugs. The entire process was genetic fishing, not chemical synthesis.
Then came Neupogen (boosting white blood cells so chemotherapy patients could tolerate higher doses), Enbrel (autoimmune). Then, in the 2010s, Repatha (PCSK9 inhibitor for lowering cholesterol) — and today, bispecific antibodies and next-generation obesity therapies. Amgen has never left the few-mile radius around that original AMGen building — but it is now the largest independent biotech company in the world.

Three things:
First thing: Don't synthesize drugs — "print" proteins using mammalian cells. This is the paradigm shift from "chemist" to "biomanufacturing."
Second thing: Fishing for a single gene may take years — but once caught, E. coli and CHO cells will replicate it infinitely. Biologic drugs are "developed once, produced forever by cells."
Third thing: Amgen has always insisted on staying independent — never merging with any major traditional pharma company. An independent biotech company can exist and outlive the old-line pharmaceutical giants.

← Previous Contents Next →
Chapter 60 ATT

March 10, 1876, Boston. Alexander Graham Bell sat in his rented laboratory and spilled a bit of sulfuric acid onto his leg. He shouted a sentence into a strange metal device — not to talk to anyone. He was calling for his assistant — "Mr. Watson — come here — I need you!" Watson was not in that room. He was in another room at the end of the hallway. But he heard every word clearly through the receiver of that device. This was the first time in human history — one person's voice, not carried by air vibration but through a copper wire — was heard by another person. Later, that copper wire became AT&T — today with a market cap of $176 billion.

Not the discovery of electricity. Not the invention of the internet. Not communications technology. Not a business model innovation.
Just Bell. The son of a Scottish educator of the deaf. He spent his entire life trying to help those who couldn't hear — "hear." The reason he invented the telephone was not to do business — he was conducting an experiment on something called the "harmonic telegraph," trying to transmit multiple messages simultaneously on a single telegraph line. Then he accidentally spilled a drop of sulfuric acid from a battery onto his leg. That inadvertent cry for help — became the starting point of all human acoustic communication.
The aftermath of that cry is the most convoluted story in the history of American communications: a company that, relying on that copper wire, monopolized all telephone calls in the United States for nearly 100 years — and was then broken apart by the government with its own hands. Then those fragments spent the next four decades devouring their own descendants — and then merging back together. And Bell himself, the man in the story — after inventing the telephone, largely stopped participating in the company's management.

Bell was an inventor, not a capitalist. In the first few years after obtaining the telephone patent, he once attempted to sell all the patents as a package for $100,000 to Western Union Telegraph Company — the most powerful communications giant in America at the time. Western Union's president William Orton glanced at it and uttered a line remembered in business history — "What use could this electric toy possibly have for the company?" He didn't buy it. Then he discovered it was far too useful — because every merchant wanted one installed in their shop. Western Union was forced into a screeching U-turn, asked Thomas Edison to invent an alternative "non-Bell patent" telephone system to compete. The two then fought the first major communications patent war in human history. Bell won.
In 1877, the Bell Telephone Company was founded. In 1885, it established a subsidiary specifically for long-distance connections — the American Telephone and Telegraph Company (AT&T). By 1913, AT&T, through mergers and patents, controlled over 85% of all telephone lines in the United States — it had effectively become "the U.S. government of telephony" — if you wanted to call your neighbor, you had to go through its switchboards. The roof of its headquarters held the largest coaxial cable hub in the world.
In 1913, it voluntarily accepted the first government regulatory agreement — the "Kingsbury Commitment" — pledging to stop swallowing independent telephone companies and to allow them access to the long-distance network. This move temporarily averted direct breakup. Then, for the next 70 years, AT&T operated as a regulated monopoly — but during that same period, it gave birth to the greatest industrial laboratory in human history: Bell Labs.
Bell Labs invented the transistor, radio astronomy, the UNIX operating system, the C programming language, the laser, information theory (Shannon), the solar cell, and the charge-coupled device (CCD, the eye of the digital camera). The transistor is one of the most important physical inventions of the 20th century. The transistor, UNIX, and C still simultaneously power your phone, the internet, and artificial intelligence systems.
So a portion of AT&T's monopoly profits — was fed back into building a "temple of science" virtually unmatched in the history of human knowledge.

In 1984, the U.S. Department of Justice won its antitrust lawsuit — AT&T was forcibly broken up. Long-distance communications stayed with AT&T; local communications were carved into seven "Baby Bells." Then, from the 1990s through the 2000s, these Baby Bells re-merged with each other — and were re-swallowed by AT&T itself. Finally, SBC Communications (a Baby Bell) bought the remnants of AT&T's long-distance division — and chose to keep using the AT&T name. So today's AT&T — is a "descendant company" among its own grandchildren wearing its ancestor's surname and trademark.
Bell himself, in this story — after 1881, never attended another board meeting. He poured large sums into deaf education, founded the journal Science (which is still published to this day) and the National Geographic Society, and funded early aerospace experiments. He insisted his entire life on not installing a telephone in his private study — because he felt the ringing would interfere with thinking.

Three things.
First thing: The one who ultimately controls communications is not the first person to get the patent — it's the one who controls the "wire in the middle."
Early AT&T didn't make its money selling telephone sets — it made its money owning the telephone lines and charging a monthly fee for them. The long-distance switchboard was a toll booth; the local exchange was "the one utility you couldn't choose."
In your industry, what is the "wire in the middle" — hardware, platform, or certification? Who owns it?
Second thing: Monopoly profits can destroy a company — or incidentally give birth to the greatest invention laboratory in human history.
Bell Labs was fed by monopoly profits — but it also invented half the foundational technologies of the modern world. So "monopoly" itself is not an absolute evil — the key is where the excess profits it generates are directed.
Are your excess profits — "feeding a future"?
Third thing: After the giant is broken apart — the fragment that most resembles the ancestor will reclaim the ancestor's surname.
AT&T died. Then SBC bought it — and put on its name like a garment. A brand lives longer than a balance sheet. Companies can perish — but a trademark can be fought over by all descendants like a crown.
Your brand right now — 100 years from now — who will want to wear it?

Bell spent his whole life speaking sign language with his deaf mother. He invented the most important voice communications tool in the world — and the person he loved most would never, in her entire life, hear his telephone ring.

← Previous Contents Next →
Chapter 61 Arista-Networks

In 2004, a former Cisco engineer named Andy Bechtolsheim founded Arista in Palo Alto. While building data centers for Google and Microsoft, he discovered a fatal crack that Cisco had ignored: the existing network switch operating system was a mountain of patches layered on top of 1980s code — and cloud data centers needed a "programmable, Linux-like switch." So he built an Extensible Operating System — EOS. Today Arista has a market cap of $175.4 billion.

Not a router revolution. Not chips. Not fiber optics. Not protocol invention.
It was Andy — the man who wrote Google its first check at Stanford (Sun's co-founder). His core insight was brutally simple: a public cloud data center is not a "bigger enterprise rack" — it is an entirely new species, and a new species needs a new neural network. He wrote the entire Linux foundation layer for Arista EOS from scratch — and every top cloud company became his customer.

Three things.
The first thing: When all your traditional competitors' software is a "geological layer cake of history" — rewriting from zero is the only moat.
The second thing: Cisco ignored the "programmability" variable while dominating the enterprise switch market. Big companies always overlook "making their most profitable product programmable" — and then get devoured through that crack.
The third thing: Sell to customers who think only about speed — your system doesn't need to convince the entire IT department, it just needs to be faster than every other system at 40G.

← Previous Contents Next →
Chapter 62 Palo-Alto-Networks

In 2005, an Israeli-born engineer named Nir Zuk drew a diagram in Sequoia Capital's Sand Hill Road office — he said a firewall is not about "ports." Real security must know whether every data packet carries Facebook or Gmail or malware. His former colleague, the founder of Check Point, told him he was wasting his time. Today Palo Alto Networks has a market cap of $174.3 billion, and enterprise security architects worldwide use his next-generation firewall standard.

Not antivirus. Not traditional perimeter security. Not passwords and tokens.
It was "App-ID" — Zuk insisted that every flow of traffic must be identified at Layer 7, the application layer, not merely checked by IP and port. Most firewalls at the time only looked at "where it came from, where it's going," blind to application content. He didn't just improve the firewall — he replaced the definition of a firewall: from "gatekeeper" to "customs inspector for every packet of information." Then he extended this entire logic to the cloud and the endpoint — creating what is called "platformized security" — network security, endpoint security, cloud security, and Security Operations Centers (SOC).

Three things.
The first thing: The previous generation of firewalls had no idea what application they were looking at — he issued an ID card to every application at Layer 7.
The second thing: The security industry sells the degree to which an admin can't sleep at night — not a feature checklist. The more you can prove "you know what's in every packet," the less the admin loses sleep.
The third thing: When your former employer is the industry standard — and you discover its biggest weakness — you yourself become the antidote to that weakness.

← Previous Contents Next →
Chapter 63 Thermo-Fisher---赛默飞

In 1903, Chicago. An instrument salesman named C.G. Fisher opened a scientific instrument company in a downtown attic. He sold only one thing: a precision analytical balance — capable of weighing to one ten-thousandth of a gram for university chemistry departments. Over a hundred years later, Fisher Scientific merged with Thermo Electron in 2006 to become Thermo Fisher Scientific. Today its market cap is $171.1 billion. It is not a pharmaceutical company — it is "the company that provides tools and reagents to every drug manufacturer, laboratory, and CDC."

Not drug discovery. Not diagnostic breakthroughs. Not biotechnology.
It was "the Walmart for scientists — but what it sells is not cheap goods, it's precision." In the first month of the COVID-19 pandemic, Thermo Fisher was one of the world's largest suppliers of PCR testing reagents and equipment. In every mass spectrometer, chromatograph, electron microscope, and gene sequencer you will ever use — the Thermo Fisher label is on the back of the machine.

Three things.
The first thing: Don't make drugs — make the common tool layer for every drug company and testing institution. The tool layer is a one-time R&D capital expense, but consumables (reagents, columns, plates) are recurring profit forever.
The second thing: The merger wasn't about cutting costs — it was about putting "lab hardware + reagents + services" into the same sales representative's briefcase.
The third thing: Become infrastructure in every invisible place — Thermo Fisher is not in the consumer's field of vision, but it stands behind the expiration date of every new drug and vaccine humanity ever launches.

← Previous Contents Next →
Chapter 64 Western-Digital

In 1970, Santa Ana, California. An engineer named Alvin B. Phillips founded Western Digital — initially making semiconductor testing equipment. Then it moved into storage controllers, and then into hard drives. After acquiring SanDisk for $19 billion in 2016, it became a dual-engine of HDD + NAND flash. Today WDC has a market cap of $168.4 billion. The photos on your phone might be backed up to iCloud — and iCloud's physical substrate almost certainly includes Western Digital hard drives.

Not phones. Not cloud services. Not SSDs. Not data center solutions.
It was "pressing the physical substrate of digital memory — rotation speed, magnetic heads, floating gates — down to the lowest cost per gigabyte." HDDs are far from dead — in data center cold storage, video archiving, and large model training datasets, HDD's cost per terabyte is still one-sixth that of NAND.

Three things.
The first thing: Storage is not "putting things in boxes" — it is "memory preserving the entire world's data at the lowest possible cost, ready to be reawakened."
The second thing: HDD + Flash dual engines — letting it capture both the long-term dividends of cloud archiving and the speed demands of PCIe SSDs.
The third thing: Data is never truly deleted — anything meant to be "kept forever" will eventually land on an HDD.

← Previous Contents Next →
Chapter 65 Gilead-Sciences---吉利德科学

In 1987, a doctor named Michael Riordan founded a small biotech company in Foster City, in the San Francisco Bay Area. On a whiteboard he wrote three words: "Cure. Not manage." — he was taking on viral hepatitis and AIDS. Years later, Gilead's sofosbuvir (Sovaldi) became in 2013 a "functional cure" for hepatitis C — capable of clearing the virus in over 90% of chronic HCV cases within 12 weeks. To permanently expel a human virus with a single oral pill — that is an extraordinarily rare moment in the history of medicine. Today Gilead has a market cap of $168.0 billion.

Not a traditional pharmaceutical company. Not broad-spectrum antibiotics. Not symptom management. Not vaccines.
It was "staring down the hardest viruses — HIV and HCV — and making drugs that can cure them." In the HIV field, Gilead developed single-tablet complete regimens capable of suppressing HIV viral loads (the Truvada, Biktarvy series) — allowing those infected to live a normal lifespan, a normal life.
Hepatitis B — not yet cured. But Gilead has been pushing in that direction day and night. And then hepatitis C — truly, functionally cured. And then it kept throwing punches in CAR-T cell therapy for tumors and the COVID antiviral remdesivir.

Three things.
The first thing: Write "cure" at the starting line of the company — not "manage," "control," "delay." Then every employee knows what they are doing every single day.
The second thing: Curing a virus means eliminating your own market — and Gilead actually did it. After curing hepatitis C, HCV division revenue plummeted — and then it had to find the next uncured virus.
The third thing: The single-tablet HIV regimen turned infection from a death sentence into "one pill a day for a chronic condition." That changed global public health.

← Previous Contents Next →
Chapter 66 BlackRock---贝莱德

In 1988, Larry Fink founded Blackstone Financial Management in a small office in Manhattan, New York. His first loss was enormous — during his time at First Boston, an investment portfolio had lost clients $100 million due to an interest rate prediction error. He nearly destroyed his own career. So he decided: no more betting on direction. He would build a machine that "quantifies every risk using mathematics" — putting all unknowns into a system called Aladdin. Today BlackRock manages over $10 trillion in assets, with a market cap of $167.3 billion.

Not a stock-picking genius. Not macro forecasting. Not a hedge fund. Not private equity.
It was Aladdin — the Risk Management Platform. This is BlackRock's heart — a real-time risk assessment and simulation engine that integrates millions of assets globally. Central banks, sovereign wealth funds, and pension systems all lease Aladdin to manage their own investment risk. BlackRock is not only the world's largest asset manager — it is part of the global financial system's risk operating system.

In the 2000s, Larry Fink began the tradition of his annual CEO letter — every year he writes to the CEOs of every publicly listed company, telling them: "Beyond profits, pay attention to climate change and stakeholders." This has been criticized by some as political posturing, and seen by others as the direction of 21st-century capital markets. Fink's response: "Risk is risk. If rising sea levels flood your supply chain — that's not politics. That's asset impairment."

Three things.
The first thing: From the burn of losing $100 million of his own clients' money grew a risk management engine.
The second thing: Aladdin isn't selling software — it's "BlackRock renting out its own operational brain to other asset managers." This turned a cost center into a profit center.
The third thing: Make the CEO letter an annual public event — embedding one asset management firm's worldview into the opening statement of the global capital allocation conversation.

← Previous Contents Next →
Chapter 67 Corning---康宁

In 1851, a merchant named Amory Houghton bought a bankrupt glass factory in Corning, New York. He used the local quartz sand — among the purest natural silica on Earth — to begin making red glass lenses for railway signal lamps. Then that glass factory became the greatest specialty glass innovation factory on the planet. Today Corning has a market cap of $167.3 billion. Its Gorilla Glass on your phone can survive countless drops from your pocket onto concrete. Its optical fiber threads transmit gigabytes of internet data per second beneath your home. And Corning's original lamp lenses — are still in production.

Not consumer electronics. Not telecommunications. Not pharmaceuticals. Not energy.
It was glass. But Corning's glass has spent its entire life doing one thing — "domesticating a transparent material to meet every new physical demand." Edison called on Corning to make molds for his incandescent light bulbs. After World War II, Corning supplied television picture tube casings. In 1970, Corning physicists pulled the first low-loss optical fiber from quartz glass — and then global broadband networks became possible. In 2007, Steve Jobs called Corning's CEO and asked him to take a thin aluminosilicate glass from the lab — Gorilla Glass — and pull it to cover the front of the first iPhone, within weeks. And later, every smartphone.

Three things.
The first thing: One hundred and seventy years, one material — glass — but every generation gives it a new physical function.
The second thing: Jobs's phone call came just weeks before mass production. Corning's answer was "Yes" — not because of luck, but because its labs already held a hundred "glass formulas that hadn't yet found their commercial use," waiting for a demand to knock on the door.
The third thing: Basic materials innovation cycles are measured in decades — but once a use is found, it can define the hardware morphology of an entire century.

← Previous Contents Next →
Chapter 68 TJX-Companies

In 1976, Bernard Cammarata opened the first T.J. Maxx in a Boston suburb. He sold leftover inventory from other department stores — last season's clothes, end-of-season sizing irregularities, excess production from brand manufacturers. He bought low and sold under the same brand names but at 70% off. Department stores called him a "flea market for discounted brands." Today TJX has a market cap of $165.1 billion and is the world's largest off-price brand retailer. In every recession, its sales rise counter-cyclically — because when the middle class tightens its belt, it retreats from Fifth Avenue to the aisles of T.J. Maxx.

Not building its own brands. Not fashion design. Not high-end department stores.
It was "brand inventory arbitrage" — using other people's brand value to sell at low prices. TJX has never designed a single garment, never produced a single home goods item. But it has one of the world's largest buying teams — combing through surplus inventory, canceled orders, and end-of-season returns from high-end brands to extract every good item that can sit under a "deeply discounted" label. And it offers a "treasure hunt experience" — every time a shopper walks in, they don't know what brand they'll find today. That random reward mechanism happens to be one of the strongest addiction loops there is.

Three things.
The first thing: Automatically expands during recessions — the harder the economy, the more consumers "hunt for discounted brands."
The second thing: Produces no products, only arbitrages the pricing gap on branded inventory. This is replacing "product rights" with "buying rights."
The third thing: The "treasure hunt" randomness brings shoppers back again and again — this turns shopping into a gambling-like dopamine cycle.

← Previous Contents Next →
Chapter 69 AppLovin

In 2012, Palo Alto. Adam Foroughi founded a mobile ad platform called AppLovin. Its logic was the opposite of everyone else's — it didn't help brands place ads, it helped app developers recommend users to each other. One mobile game ad promoting another mobile game — with machine learning automatically optimizing the placement. Then he built his own game studios, used his own platform to promote his own games, and used the data feedback loop to feed himself into becoming the company on Earth that understands best "how to get a user to download an app." Today AppLovin has a market cap of $164.0 billion.

Not brand advertising. Not a social platform. Not search placement. Not content marketing.
It was "using machine learning to press the cost of 'installing an app' down to a level where no competitor can survive." Foroughi has remained hidden from public view for years — AppLovin's AI placement engine, AXON, when it comes to optimizing app install costs, is considered by many developers within the mobile ecosystem to be more efficient than Google and Meta.

Three things.
The first thing: Create your own "demand" — develop games in-house, promote them on your own platform, form a data closed loop. External clients' data fattens the engine, in-house games contribute the profits.
The second thing: Don't compete with Google and Facebook on brand advertising — only slice off the "app install" vertical behavior. Narrow to the extreme becomes wide.
The third thing: The true nuclear weapon of machine learning in advertising is not creative — it's "the cost function for acquiring the right user at the right time at the right price."

← Previous Contents Next →
Chapter 70 Deere---迪尔

In 1837, the Illinois prairie. A blacksmith named John Deere moved here from Vermont. The local pioneer farmers used cast-iron plows that worked in the sandy soils of the East — but the Midwest was thick, sticky chernozem. The moment an iron plow went in, it was like sinking into asphalt. Every few steps, a farmer had to stop and scrape the mud off with a wooden stick. Deere took a discarded saw blade from a lumber mill, cut it into a mirror-polished curved surface — and made the world's first "self-scouring steel plow." Today his company has a market cap of $157.1 billion, and the green Deere tractor is one of the largest agricultural machines in the world.

Not the inventor of the tractor. Not the patent holder for the internal combustion engine. Not the inventor of GPS.
It was John Deere. Blacksmith. He didn't build tractors — he built a plow. But that "self-scouring" physics principle was carved into Deere's DNA from that moment on. One hundred and seventy years later, the cab of a Deere 9R Series tractor is full of touchscreens — sensors measuring soil organic matter in real time, seed spacing, the variable rates required for spraying. But the steel plow beneath is still there — the principle unchanged — only now it's a precision GPS-guided cutting wheel instead of a single saw blade.

Three things.
The first thing: Don't build your factory in the East — move to the very center of the prairie clay. Test the self-scouring curve of the prototype on the toughest local soil, over and over, until nothing sticks.
The second thing: He didn't ask "what do farmers want." He crouched in the mud for hours — and learned exactly how a plow behaves in every inch of soil. Have you logged enough "mud-crouching" time?
The third thing: From plow to autonomous tractor — but the shell has always been that same green. A brand color is worth more than advertising.

← Previous Contents Next →
Chapter 71 Uber

Winter 2008, Paris. Two tech entrepreneurs stood beneath the Eiffel Tower for 45 minutes, unable to hail a single taxi. Snow was falling. One looked at the other and said: "Wouldn't it be cool if you could press a button and a car just showed up?" The other said: that's called a limousine service, and it's insanely expensive. The first one said: "What if everyone could?" Today Uber has a market cap of $155.8 billion. It doesn't own a single car. It is the largest "asset-less" mobility company on Earth.

Not a taxi company. Not automotive innovation. Not self-driving (at the start). Not driver employment.
It was Travis Kalanick and Garrett Camp. Two serial entrepreneurs — Kalanick had been run nearly into bankruptcy by his first company, Red Swoosh, then turned it around when Akamai acquired it for $19 million. Then they built a black car hailing app you could operate with a single finger — UberCab. The first version had exactly three black Lincoln Town Cars, which Kalanick personally flagged down on the streets of San Francisco. He interviewed every single driver himself.
Nine years later, Uber was providing rides for tens of millions of people worldwide every day — while simultaneously being sued by governments, struck against by taxi drivers, and coup'd by its investors — nearly burning itself to ash. Then Dara Khosrowshahi took over, pinned this berserk speed machine to the ground — and turned it into a company that could actually turn a profit.
Let me tell this story in two parts. When I'm done, we'll talk about why "insane speed" can ignite an entire industry — and also nearly extinguish the match that lit it.

Part One: Lighting the powder keg.
Kalanick is a natural-born antagonist. He plays squash, bridge, business warfare — he loves the fight itself. In 1998 he dropped out of UCLA to build Scout — a P2P file-sharing company, sued into bankruptcy threat by the RIAA. Red Swoosh — his second company — betrayed by investors, sabotaged by his co-founder, at his lowest point his bank balance was negative, surviving on a few thousand dollars from his mother to handle IRS penalties. Then in 2007, Akamai bought Red Swoosh — and he crawled out of debt in a single stroke.
That experience left two deep marks: One, a primal distrust of every "vested interest" — government, unions, legacy industries. Two, a permanent fear that cash could run out — so you had to run as fast as physically possible.

In 2009, UberCab's first year running in San Francisco — just three cars. Kalanick personally recruited the first wave of customers on Twitter. In 2010, they received a "cease and desist" order from the San Francisco Municipal Transportation Agency and the California Public Utilities Commission — they were not a taxi company and held no road transport license. Kalanick did not stop. He changed the company name from UberCab to Uber — dropping the "Cab" — and kept operating. He also hired digital strategists from Obama's campaign to launch demonstrations and on-demand lobbying.
From 2012 to 2016, Uber fought over a hundred open battles with regulators worldwide — its cars intercepted on the streets of Paris by taxi drivers, its license revoked by Transport for London, investigated simultaneously by Beijing and Brussels. Uber's strategy was almost always the same: launch first — get the public to fall in love — use email and in-app pop-ups to turn citizens into an automated lobbying weapon — then negotiate a legal framework with the government afterward.
This "go in, burn, then force a conversation" expansion placed enormous physical and psychological stress on everyone who worked with Kalanick. But it also allowed Uber to cover hundreds of cities globally within five years.

Part Two: The fire nearly killed the person who lit it.
2017 — Uber was hit by a cascade of devastating crises: former employees exposed a culture of sexual harassment, Google's Waymo was suing Uber for stealing its self-driving trade secrets, #DeleteUber erupted during the Trump travel ban, and the dashcam video of Kalanick losing his temper at an Uber driver went viral across the entire internet. Then came the boardroom coup — led by Benchmark and multiple major investors — Kalanick was forced out as CEO.
Dara Khosrowshahi — former CEO of Expedia — took over this giant infant, still screaming in the flames of public relations and regulatory fire.
He did several profoundly "un-Uber" things: publicly apologized to Transport for London and applied for a legal license (Uber's previous approach in London had been defiance). He rewrote Uber's toxic "win at all costs" internal values — internally they called it "We do the right thing. Period." He also took the company public in 2019. After losing money for over a decade, in 2023–2024 Uber finally began generating sustained positive free cash flow.
Then he discovered that Uber's real profit wasn't in moving people — it was in moving things. Food delivery (Uber Eats), freight, grocery delivery — this "logistics network on wheels" proved more predictable, higher frequency, and easier to automate than carrying passengers.
Uber became "the AWS of urban on-demand transport" — not just a rides app.

Three things.
The first thing: The cracks in old rules are the most valuable entry points.
Every city taxi system in the world is "medallion caps + government-set pricing + street hailing" — a crack that hasn't changed in 100 years. Uber replaced every piece of that old crack with three pure software variables: GPS, dynamic pricing, and two-way ratings. Kalanick wasn't the first person to see this crack — but he was the person who ran through it fastest.
Where is the 100-year-old "old crack" in your industry?
The second thing: Speed is a weapon, and a time bomb.
"Move fast and break things" can fling your competitors out of the visible frame — but it can also bury the women and minorities inside your own company under the rubble. Dara's first task after taking over wasn't growth — it was stopping the bleeding, apologizing, and rebuilding the rules.
Can your internal culture, your diversity, and your regulatory foundation withstand triple-speed expansion? If not — you are building a bridge that will collapse.
The third thing: Food delivery and the data layer often have higher profitability than the "original star product."
Uber poured every resource into passenger rides. Then Uber Eats became its most stable profit engine, because food delivery route structures and time windows are easier to predict, and large volumes of deliveries can be pre-orchestrated by algorithms. That most overlooked "side hustle" inside your company — is it actually your most profitable future product?

Uber didn't invent the car. Didn't invent the taxi. Didn't invent food delivery. It invented this: "anyone with wheels and idle time can become a node in an on-demand transportation network." And then it dismantled a wall built from a hundred years of licensing bricks — piece by piece, with cell signals, dynamic pricing, and a rating system.

← Previous Contents Next →
Chapter 72 Charles-Schwab---嘉信理财

On May 1, 1975 — the very day the "May Day" deregulation took effect — a 34-year-old stockbroker named Chuck Schwab registered a discount brokerage firm called Charles Schwab in San Francisco. His office had a single folding chair, a metal desk, and a telephone. He offered no analyst reports, no investment advice. He did only one thing: help people buy and sell stocks at rock-bottom commissions. Wall Street laughed at him — "You won't even give advice. Who would come?" Today Charles Schwab has a market cap of $154.7 billion.

Not an investment bank. Not a fund. Not quantitative trading. Not a tech platform.
It was Schwab. A Stanford MBA who had struggled academically due to dyslexia. He discovered a secret that was hiding in plain sight — in traditional "full-service" brokerage commissions, most of the money went to something the client never actually used: so-called "investment advice." And that advice was often wrong. Schwab slashed commissions from hundreds of dollars to tens of dollars — no advice, execution only. Then thousands upon thousands of individual investors discovered: no advice — making their own judgments — was cheaper, and the returns were no worse. Then he pushed this discount model through the internet era — becoming the largest online retail brokerage. Then he launched banking, an ETF platform (one of Schwab's ETFs now carries the lowest fee rate in the world), and finally in 2020 acquired TD Ameritrade for $26 billion.

Three things.
The first thing: When everyone else was packaging "advice" as justification for higher fees — he cut the advice and pounded the fees to the floor.
The second thing: Dyslexia made him bad at writing long reports — so he decided not to write them. His system runs on "giving users accurate information and letting them decide for themselves." His personal limitation became the entire company's product philosophy.
The third thing: Always devour "zero commission" at the lowest point of the interest rate cycle — then earn the spread on uninvested cash sitting in accounts. He doesn't make money on commissions — he makes money on the spread.

← Previous Contents Next →
Chapter 73 Uber

On a freezing winter night in 2008, beneath the Eiffel Tower in Paris. Garrett Camp and Travis Kalanick had just left a party at a tech conference — it was raining outside. They each stood in the middle of the road trying to hail a taxi for 30 minutes. Camp turned to Kalanick and said: "My dream right now is — pressing one button on my phone and a car just comes." Kalanick had just sold his previous company, Red Swoosh, and was in a state of "financially free but irritable." He said: "Then why doesn't it exist?" And Uber was pushed into the world. Today Uber has a market cap of $155.0 billion.

Not the sharing economy. Not gig labor legislation. Not self-driving. Not taxi unions.
It was two people who couldn't get a car in the freezing rain. Camp had just sold StumbleUpon to eBay and was in Paris for the LeWeb tech conference. After returning to San Francisco, he first registered the domain UberCab.com — note that the name still had "Cab" in it — then bought a few Mercedes S550s, hired drivers, and operated only within San Francisco city limits. That earliest version of Uber was not "ride-sharing" — it was a mobile private limousine fleet. A single ride cost roughly 1.5 times a normal taxi — but because you could summon one with a single press on your phone, and the driver wouldn't invent an excuse to refuse the fare, it instantly won word-of-mouth across the San Francisco tech scene.

Then Kalanick took over as CEO — and turned Uber from a "limousine app" into "alternative transport for anyone in any city." His method came down to three words: get in first. City by city — pulling drivers and passengers onto the platform before the local government even knew what a "ride-hailing service" was. By the time regulators caught on and moved to ban it, the platform already had millions of users.

In 2015, the Paris city government banned Uber's low-cost service UberPop — Kalanick's response was to dispatch a lobbying team while refusing to cancel the service. In 2017, Transport for London revoked Uber's operating license — Kalanick's response was to sue while simultaneously issuing an open letter apologizing and pledging to improve driver background checks. This "break in first, negotiate later" playbook defined the regulatory culture of the entire ride-hailing industry.

Then in 2017, Kalanick was forced out as CEO by the board due to internal culture problems and massive controversy. He later sold most of his shares — and is now working on restaurant robotics. Under Dara Khosrowshahi's leadership, Uber sheathed its claws — went public, cut unprofitable projects, sold off its operations in China, Southeast Asia, and Russia, and finally achieved annual profitability in 2023. Khosrowshahi's strategy was "disciplined scaling" — the polar opposite of Kalanick's "growth till death."

Three things.
The first thing: The best product is the one you think up in the moment you are cursing in the freezing rain.
Camp and Kalanick failing to hail a taxi in Paris at night was not a one-time experience — everyone has lived through it. But the two of them happened to be tech entrepreneurs, and they happened to have just come from a "Future of Tech" conference that very night.
That inconvenience you curse at every day — why isn't it your product?
The second thing: The "get in first" growth method — is a double-edged blade.
Uber used "do it first, negotiate later" to break into 70 countries and 10,000 cities — but also accumulated over a thousand lawsuits, billions of dollars in fines, and revoked licenses. Without that edge — you can't start the game. But with that edge — Kalanick himself was ultimately cut to death by it.
Your growth — are you running fast inside the rules, or planting seeds ahead of where the rules have yet to reach?
The third thing: The founder was driven out — the company survived, and got a little better.
After Kalanick left, Khosrowshahi turned Uber from a "cash-burning machine" into a company with positive free cash flow. This is not a negation of the founder — it is the recognition that the founding phase and the scaling phase require entirely different CEO personalities.
At your company's current stage — does it need an "entrepreneur" or an "operator"?

Garrett Camp rarely appears in the media anymore. He quietly bought a substantial portion of a Hawaiian island — not to build a hotel, but for nature conservation. He said: "On the island, I haven't used a single ride-hailing app."

← Previous Contents Next →
Chapter 74 Welltower

In 1970, Toledo, Ohio. A real estate investor named Fred Carr made a decision in the healthcare REIT space that no property developer at the time could understand — buy only hospital buildings and senior care facilities, then lease them to medical operators. For his entire life he did only one type of property: "buildings where patients go to receive care." Today Welltower has a market cap of $153.0 billion. It is one of the largest healthcare real estate investment trusts in America. It doesn't treat illness — it owns the physical space where healing happens.

Not hospital operations. Not medical care. Not insurance. Not pharmacies.
It was stripping "healthcare real estate" out from commercial real estate — making it a standalone asset class. When retail and office real estate are being hollowed out by the twin blows of e-commerce and remote work — hospitals, clinics, and nursing homes are mandatory physical spaces that cannot be replaced online. Welltower holds a vast portfolio of these assets in states with the most severe aging demographics and the densest Sun Belt populations. Its logic has never been "next quarter's rent" — it's "the elderly population will double over the next 30 years."

Three things.
The first thing: Own only one type of physical space that can't go online — healthcare and senior living — then wait for demographics to do the work for you.
The second thing: REITs allow ordinary people to indirectly own hospitals and nursing homes — this turns "aging" into an asset class that can be traded on capital markets.
The third thing: Don't touch retail, don't touch office — every time the market hypes those assets, maintain absolute non-contact. Focus = immunity to every short-term fever.

← Previous Contents Next →
Chapter 75 Intuitive-Surgical

In 1995, inside a lab at the Stanford Research Institute, a group of scientists was investigating "how to let a surgeon operate without putting their own hands directly inside a patient's body." Then Intuitive Surgical's founder, Dr. Frederic Moll, turned that into a machine called the da Vinci Surgical System — it reproduces the surgeon's minuscule, tremor-free finger movements inside the patient's body through four robotic arms, while magnified 3D high-definition imaging lets the surgeon see microscopic nerves invisible to the naked eye. Today Intuitive Surgical has a market cap of $152.2 billion.

Not a new drug. Not genetics. Not diagnostics. Not assistive devices.
It was "putting the surgeon inside a cockpit of microscopes and joysticks." Dr. Moll was originally a cardiac surgeon. The reason he built a robot — was that heart surgery required sawing open the sternum, inflicting massive trauma. With da Vinci, you only need a few small incisions. Then urology, gynecology, and head and neck surgery all replicated the same logic.

Three things.
The first thing: It's not a robot doing the surgery — it's a surgeon remotely operating a robot to do the surgery. Human judgment always stays in the highest decision-making seat.
The second thing: Every da Vinci is a platform for repeatedly consuming disposable instruments — the profit doesn't come from selling that machine, it comes from the forceps, scissors, and energy instruments that must be replaced for every single surgery. This is the razor-and-blade model.
The third thing: Moll himself was a cardiac surgeon — he was solving the physical limits of his own hands.

← Previous Contents Next →
Chapter 76 Dell

In 1984, a University of Texas at Austin dorm room. A 19-year-old named Michael Dell began selling upgraded IBM PC compatibles he assembled himself — cheaper than stores, built to the customer's specifications. His roommate hid his boxes of parts in the shower. Today Dell Technologies has a market cap of $151.2 billion. His core philosophy — "direct sales, build-to-order, zero inventory" — surpassed both Compaq and IBM, one after the other, in the PC era.

Not patents. Not chips. Not operating systems. Not software.
It was "build-to-order" — completely flipping the PC supply chain from the front-end model of "the manufacturer guesses what you'll buy and stacks it in a warehouse" to "you tell me the configuration you want, and I'll assemble and ship it in two days." At its peak, Dell's cash conversion cycle was negative thirty-plus days — the customer paid first, and only then did Dell purchase components. Using other people's money to fund its own inventory. This is one of the most cash-efficient manufacturing models in existence.

Three things.
The first thing: It's not that the computers were better than anyone else's — it's that "the wait time before you got your computer" was shorter than anyone else's, and "the configuration is exactly what you wanted."
The second thing: Negative cash conversion cycle — you don't need to pay suppliers until after you've received customer payment. This beat every PC manufacturer using supply chain math.
The third thing: The 2016 Dell-EMC merger — expanding from PCs into data storage and cloud infrastructure — transforming itself from a box vendor into an enterprise IT platform.

← Previous Contents Next →
Chapter 77 Intuitive-Surgical

In 1986, a basement laboratory at the Stanford Research Institute (SRI). A microelectronics engineer named Dr. Phil Green was conducting microsurgery experiments under an electron microscope using a needle so fine it was nearly invisible to the naked eye. His question wasn't "can I cut tissue" — of course he could. His question was: "The human hand trembles. My hand trembles 0.3 millimeters under the scope. On an artery 0.5 millimeters in diameter, that tremor means plunging the blade through the vessel wall." So he made a decision: don't train the hand not to tremble — build a machine that filters out the tremor. That decision — was the logical starting point for every surgical robot in the world. It later became Intuitive Surgical — today with a market cap of $152.0 billion.

Not artificial intelligence. Not microelectronics. Not FDA approval. Not insurance reimbursement.
It was eliminating "hand tremor" — a physiological limitation that human surgeons have lived with since the age of ancient Egypt — using engineering.

In the 1980s, SRI was developing a remote surgery robot for the U.S. Army — the vision was to let a surgeon operate from the rear while performing surgery on wounded soldiers at the front. The project was called the "Green Telepresence Surgery System," funded by the military. Dr. Green and the engineering team built the first prototype capable of remotely manipulating surgical instruments while providing haptic feedback. Then the military program was terminated — and the R&D team looked at that prototype and made a judgment entirely unrelated to the military: "The biggest market for this thing isn't in a war zone — it's in every hospital operating room."

In 1995, Frederic Moll — a serial entrepreneur in medical device startups and a Stanford MBA — licensed the technology from SRI. He founded Intuitive Surgical and in 1999 launched the first-generation da Vinci Surgical System — named after Leonardo da Vinci, because the Renaissance polymath had drawn in 1495 what is recorded as the first "design for a robot" in human history.

The da Vinci's core selling points were terrifyingly complete: high-definition 3D vision — magnified 10 times; robotic arms with 7 degrees of freedom, more than the human wrist — capable of 540 degrees of rotation; software that filters out all hand tremor completely; and the ability to operate remotely. The first installed da Vinci cost over $1 million, with an annual six-figure maintenance fee. Hospitals complained about the price — but they bought. Because after prostatectomies were performed with the robot, patient hospital stays dropped from 7 days to 1–2 days, and the blood transfusion rate fell from 30% to near zero. Economics won.

Then Intuitive Surgical locked down the market using the "razor-and-blade" model: the da Vinci system requires a console unit ($1M–$2.5M), and every single surgery must use Intuitive's disposable surgical instruments — each costing $1,500–$3,500, which automatically expire after 10 uses. The more surgeries performed, the greater the consumables revenue. As of 2025, over 10,000 da Vinci systems have been installed worldwide, completing more than 15 million surgeries — the majority being prostatectomies, gynecological procedures, and general surgery. Every 20 seconds, a da Vinci surgery is taking place somewhere on the planet.

Three things.
The first thing: It didn't design a new surgery — it eliminated the physical boundary of the human hand.
Hand tremor of 0.3 millimeters — in ordinary surgery, it's not a problem. In microvascular anastomosis or prostate cancer surgery (where the gland sits deep in the pelvic cavity surrounded by a plexus of nerves) — that tremor is the difference between "normal complications" and "postoperative incontinence." Intuitive Surgical solved just one problem, a problem at the level of physical magnitude.
In your industry — is there a physiological or physical limit that "nobody thinks can be improved" — waiting for you to eliminate it with sensors and software?
The second thing: Invented by the military — scaled by commerce.
Remote surgery robots started as a military project — because wounded soldiers on the battlefield are often thousands of kilometers from a qualified surgeon. But what Intuitive Surgical saw was this: the two largest healthcare markets in the world — the United States and Japan — both have high labor costs, and a robot could standardize a surgeon's service radius.
Is there a piece of "military technology" near you — waiting for its civilian-scale entrepreneur?
The third thing: The "razor + blade" economics of machine + consumables — works just as well inside a hospital.
The consumable cost for a single da Vinci surgery is approximately $1,500. The surgical fee the hospital charges the patient (robot-assisted) includes the insurance reimbursement portion. The robot specifically adds a "disposable 3D endoscope" and "multi-joint staplers" — which must be replaced after every use. Intuitive Surgical's consumables revenue has already surpassed the revenue from selling the machines themselves.
In your business model — is there an asset that could be turned into a "consumable"?

Frederic Moll retired from Intuitive Surgical in 2019. He is a quiet man, rarely making public statements. In the basement of his home, he set up a small machine shop — spending his retirement making handmade wooden boat models. He said: "The machine doesn't tremble. Neither does the wood — but my hands have finally started to."

← Previous Contents Next →
Chapter 78 Southern-Copper---南方铜业

In 1952, in the Arequipa Valley of southern Peru, the American Asarco copper company and the Peruvian government jointly opened two colossal open-pit copper mines: Cuajone and Toquepala. At first this was a traditional contract between a handful of multinational companies and state-owned entities. Then Southern Copper, through mergers and multi-country operations across Latin America, became one of the world's largest copper producers — controlling approximately 6% of global copper output. Today its market cap is $149.5 billion. Its copper — lying underground for two hundred million years — eventually becomes the electrical wiring in every city on Earth, the electromagnetic coils in every electric vehicle, the conductive substrate for every wind turbine and solar power station.

Not a tech startup. Not a financial model. Not futures trading.
It was copper. In the mountains of Chile, Peru, and Mexico. This is not "a single founder's story" — it is "the collective long-cycle investment of generations of mining engineers and geologists." A copper deposit, from discovery to production, typically requires ten to twenty years. Southern Copper's expansion is not measured in quarters — it is measured in centuries.
Copper's electrical conductivity makes it the foundational metal of decarbonization and electrification — an electric vehicle requires three to four times the copper of a traditional internal combustion engine vehicle. Meanwhile, new copper mines globally are becoming harder and slower to permit. Its intrinsic value is not the copper price of the last quarter — it is "the irreplaceable physical element in the world's future conductive infrastructure."

Three things.
The first thing: Resources are not created — they are discovered, on geological timescales. Mining is long-term capital allocation with geological history.
The second thing: Copper's irreplaceability — nowhere on the entire periodic table is there a second element whose combination of conductivity + ductility + cost comes anywhere near copper.
The third thing: A production timeline of over a decade means — when copper demand explodes, supply elasticity is extremely low. Those who positioned first lie atop the strata and wait.

← Previous Contents Next →
Chapter 79 Pfizer

In 1849, Brooklyn. Two German cousins — Charles Pfizer and Charles Erhart — opened a chemical workshop on the second floor of a red-brick building. Their first product was a white powder that tasted like sugar but could kill intestinal parasites. They mixed it with toffee — creating the world's first "medicine that tasted good." Today Pfizer has a market cap of $147.6 billion. In 2020, it rewrote the history of pandemics with an mRNA vaccine. And its starting point was two German immigrants who could barely speak English — and a copper pot filled with sugar and worm-killing compounds.

Not a medical breakthrough. Not basic research. Not public health investment. Not government contracts.
It was Charles Pfizer. A confectioner's apprentice from Ludwigsburg. He took medicine — which in 1849 tasted like a mixture of veterinary sulfur and turpentine — and coated it with sugar and fruit flavoring. He didn't discover a new chemical compound. He discovered that "if medicine tastes good, patients will actually take it" — a fact so blindingly obvious that every pharmacist of the era had missed it. And then Pfizer turned "palatability" into one of the most durable competitive advantages in the history of human pharmaceuticals.

Charles Pfizer was born in 1824 in Württemberg, Germany. As a teenager, he apprenticed under a distant relative as a confectioner — learning how to heat sugar to specific temperatures and cool it into candy shells of varying hardness. What he brought to America was not just sugar — it was a technical intuition for temperature control and crystallization. His cousin Charles Erhart was a chemist — the two moved to Williamsburg, Brooklyn, borrowed $2,500 from a business associate of Pfizer's father in America, and opened a fine chemicals company called Charles Pfizer & Company.
The first product was santonin — a deworming agent extracted from wormwood. Extremely bitter. Pfizer fused it with toffee and shaped it into small conical sugar lozenges — the shape reminiscent of something out of Harry Potter, a medicinal candy you could swallow. This was the first time in pharmaceutical history that "patients would voluntarily take their own medicine" — children thought they were candy and snatched them out of their parents' hands. This santonin lozenge alone fed Pfizer for its first decade-plus.
Then the Civil War came. The Union Army needed massive quantities of painkillers and antiseptics. Pfizer scaled up its chemical synthesis of tartaric acid and iodides — bypassing tariffs on borax and camphor that had previously only been imported from Britain. The war transformed Pfizer from a small workshop into one of America's core suppliers of pharmaceutical raw materials.

In 1928, Alexander Fleming discovered penicillin — but no one could mass-produce it through fermentation. After Pearl Harbor in 1941, the U.S. government demanded every pharmaceutical company find a way to mass-produce penicillin. Pfizer's chemist James Currie — a man with a peculiar passion for mold — used deep-tank fermentation technology he had discovered in a converted old ice cream factory in Brooklyn, New York, to boost the yield of penicillin production by several orders of magnitude. On D-Day in 1944 — most of the penicillin used by wounded Allied soldiers had been brewed in Pfizer's tanks.
Then Pfizer used the fermentation and chemical capacity it had accumulated from penicillin — and in the 1950s began developing its own new drugs. This produced early antibiotics like Terramycin — beginning the shift from a "contract active pharmaceutical ingredient manufacturer" to a "proprietary drug company."
The 1990s and 2000s — Pfizer executed the most aggressive wave of patent-drug acquisitions in history, bringing Lipitor (one of the best-selling lipid-lowering drugs of all time), Viagra, pregabalin, and pneumococcal conjugate vaccines into a single house. Then in 2018–2024 — Patheon, Array BioPharma, and the extraordinarily expensive Seagen (acquired in 2023 for $43 billion, an antibody-drug conjugate company) — it placed all its bets on oncology and rare diseases.

April 2020, the COVID-19 global pandemic. Pfizer CEO Albert Bourla made a decision that provoked extreme internal controversy — take zero upfront government R&D funding, fast-track the mRNA vaccine with Germany's BioNTech, and shoulder the full risk of failure entirely on their own. He later explained: "I didn't want Pfizer's scientists slowed down by a single second because of government budget cycles and reporting paperwork. If we lose — I'll pay for it myself. If our data is wrong — we take the blame ourselves." November 2020, the Phase III results were released — efficacy exceeded 90%. That was the beginning of the end of the pandemic. That decision redefined for the entire world how a company that began 170 years earlier with sugar-coated pills operates — not by political logic, but by the old German apprentice logic of "chemistry + speed + all-in bets."

Three things.
The first thing: Do something so blindingly obvious that nobody bothers doing it — and that is disruption.
Pfizer's first decade was just "making medicine sweet." No new molecules. But the single variable of "tastes good" allowed him to devour the majority of the santonin market. In your industry — what is the most basic "palatability" layer that everyone has forgotten to get right?
The second thing: Control fermentation — and you control the industrial foundation of the antibiotic era.
Pfizer's core breakthrough on penicillin was not biology — it was scaling up fermentation engineering. Its chemist turned deep-tank fermentation into industrial production — and those fermentation tanks could then produce other strain-based drugs for the next forty years. Can your core manufacturing step — be horizontally transferred to your next generation of products?
The third thing: Reject free money before the biggest bet — in exchange for full speed.
Pfizer refused government COVID R&D funding not out of principle — but for speed. Because taking government money comes with mandatory compliance reports and regulatory reviews, and a pandemic doesn't wait for paperwork. Do you have the courage to refuse money that "looks like free money but is a lethal deceleration to your speed"?

That red-brick building in Brooklyn from 1849 is gone now. But Pfizer keeps an old copper column salvaged from the Williamsburg factory at its New York headquarters — no words are engraved on it. Just the green patina of age.

← Previous Contents Next →
Chapter 80 Abbott-Laboratories---雅培

In 1888, Chicago. Dr. Wallace C. Abbott began manufacturing a medicine called "Abbott's Granules" in the kitchen of his own apartment — he placed alkaloids into extremely precise small-dose granules so that doctors could administer accurate doses to patients during house calls without needing a scale. He was a practicing physician, not a pharmacist — but every day he saw patients being poisoned by inaccurate dosing. Today Abbott Laboratories has a market cap of $145.8 billion, and its FreeStyle Libre continuous glucose monitor is worn on the upper arms of tens of millions of diabetics worldwide. Abbott doesn't do science fiction. It does only one thing: "precision that patients can use every day."

Not new drug discovery. Not gene therapy. Not cancer breakthroughs.
It was "diagnostics + nutrition + established medicines" — executing the least hyped parts of basic healthcare with the utmost precision. Abbott's glucose monitor uses a probe thinner than a human hair to automatically measure interstitial fluid glucose once per minute — patients never need to prick their fingers. Its infant formula Similac generates billions of dollars in annual sales, nourishing premature infants. It took the most uncool but highest-frequency things in healthcare — blood glucose, nutrition, diagnostics — and turned them into a stable cash flow that has lasted over a century.

Three things.
The first thing: The founder, Wallace Abbott, was a practicing physician — what he saw was "inaccurate dosing kills people," not "return on investment."
The second thing: Place every product inside the "few things a patient must do every day" — test glucose, feed a baby, check a heart. This is not a one-time purchase; it repeats every single day.
The third thing: The FreeStyle Libre innovation — not inventing glucose measurement, but making "painless continuous monitoring" medically normal.

← Previous Contents Next →
Chapter 81 Salesforce

In 1999, a former Oracle vice president founded a company called Salesforce from his apartment. His first investor, Larry Ellison, gave him $2 million. Then he did something that even Ellison thought went too far — he rented a plane during the Super Bowl and flew a banner over Silicon Valley: "Salesforce.com — The End of Software." After that, he hired actors to stage fake protests outside competitors' conferences, holding signs that read "Down with Software!" This was when he was at his poorest. Today, Salesforce is worth $143 billion.

Not better software. Not lower prices. Not stronger features. Not a bigger sales team.

It was Marc Benioff. A sales prodigy who became Oracle's youngest vice president at 27. He went on a three-month "sabbatical" in Hawaii, swam with dolphins for days, and then returned to San Francisco and made a decision — enterprise software should not be sold in boxes. It should be like Amazon: open a browser and use it.

He called this idea "The End of Software." The entire industry laughed at him. Then he dragged the whole enterprise software industry into the cloud era.

Let me tell his story first. After I finish, we'll talk about why the word "end" — placed in front of your old industry — might be the most dangerous and most valuable startup manifesto ever written.

Marc Benioff wrote his first game at 13 (priced at $75; he sold only a few copies). By 15, in high school, he was already writing assembly code for the Apple II. In 1984, he called Apple CEO Steve Jobs and said, "I'm a high school developer, and I want to intern at Apple." Jobs gave him a summer internship. The desk he later used was always "the black leather chair Steve Jobs once sat in" — he specifically bought that chair from Apple.

In 1986, he joined Oracle, starting from the customer service hotline. By 27, he had become the company's youngest vice president — earning over a million a year, with his own private plane and a husky named Koa.

In 1999, he took a vacation in Hawaii, swam with dolphins, and read a book on "consciousness transformation." Then he wrote an idea in a small notebook — "Software should not be installed, should not be maintained, should not be upgraded every 18 months. It should be like a utility — open it in a browser and use it, pay as you go."

After returning to San Francisco, he founded Salesforce in a vacant apartment next to his own Telegraph Hill apartment. He poured in his entire savings. Ellison later invested too — Ellison had a principle: he liked funding companies that went to war with existing giants, because win or lose, that war would weaken their mutual enemies (SAP and Microsoft). Benioff never hid his fundamental conflict with Oracle — he simply said, "Thank you for the money — but some of the old things I intend to kill belong to your family."

But the biggest problem early on wasn't the product. It was that nobody had heard of this new company. "Cloud software" and SaaS were not hot terms in 1999 — most IT departments didn't even want to discuss the concept of "putting your data on someone else's servers." Benioff's marketing budget was extremely limited.

So he launched a series of "performance art marketing stunts" — staging fake protests outside ERP conferences with "No Software" signs, getting chased off by venue security. Hauling in a horse-drawn carriage to perform a classical theater-style piece on "the death of old software" in downtown San Francisco. Renting a plane during Super Bowl weekend to fly a banner: "Salesforce.com — The End of Software."

Every traditional media outlet named it "the year's most tasteless marketing." But IT managers — the ones who maintained those clunky internal systems every single day — they remembered the name. "No Software" didn't mean no software at all. It meant not the installations, the patches, the outages they hated with every fiber of their being every day. Then in 2004, Salesforce went public — it picked CRM for its ticker — and the stock jumped 56% on its first day.

Over the next 20 years, Benioff grew the Salesforce empire — while turning a portion of its profits into one of San Francisco's largest philanthropic engines. He wrote the "1-1-1 model" into the company's charter — 1% of company equity, 1% of product given free to nonprofits, 1% of employee hours for volunteering. The 1-1-1 model was later adopted by hundreds of companies, including Google, Cisco, and Workday. Meanwhile, he bought vast amounts of land and property in San Francisco and Hawaii — and donated hundreds of millions of dollars to UCSF Children's Hospital and Hawaii's healthcare system.

In 2018, he acquired MuleSoft for $6.5 billion — expanding himself from a CRM tool into "a data translator between all cloud applications." In 2021, he acquired Slack for $27.7 billion — extending Salesforce from "managing customers" to "real-time connection between teams."

Critics called it imperial overreach. Benioff's answer: "Enterprise software is not a product suite — it is a unified data fabric inside the organization. We are not buying software. We are stitching organizations together."

Stop here for a moment.

Have you ever cursed your industry's "old way" — not privately, but by spending your most creative energy and what little money you had, at the moment the entire industry was watching — in a way they could never forget?

Benioff wasn't "marketing." He was declaring war. "The End of Software" transformed Salesforce from "another CRM vendor" into "the first signal of the old regime's overthrow." Do you have a sentence — one that pierces your industry's deepest pain point — and your product happens to be the answer to that sentence?

Three things.

First: "No Software" is a rebel insignia — not a feature list.

Salesforce didn't compete with Siebel and SAP on features. It used the visual language of "the old boxed world is about to vanish" — what it struck wasn't the CIO's logic center, but their subconscious, already brimming with rage at the old systems. What's your rebel insignia?

Second: Turn philanthropy into the source code of your company's DNA.

Benioff locked in the 1-1-1 model when Salesforce's valuation was still tiny — because he knew that once a company crosses a certain size, "doing good" becomes a slide in the marketing department. The things you truly believe in — you must write them into legal documents while you are still fragile.

Has what your company "believes in" been locked into legal and equity structures — or does it live only in the goodwill of the CEO?

Third: Always "stitch the organization" — don't pile on features.

Every new company he bought — MuleSoft stitching data, Slack stitching conversation — he was using the threads he purchased to sew together the organization's severed organs. When "stitching" becomes a product — what you are selling is no longer software; it is the organization's internal "nerve growth factor."

Is your product "selling organs" — or "sewing the nerves between organs"?

Marc Benioff now sits at the top of Salesforce Tower — the tallest building in San Francisco. Outside his office hangs a photograph — taken during his 1984 internship with Steve Jobs. On the wall next to that photograph, he wrote three words in marker — "End is beginning."

← Previous Contents Next →
Chapter 82 ConocoPhillips---康菲石油

In 1875, in Ogden, Utah, a small company called the "Continental Oil and Transportation Company" began hauling kerosene between the western states — using a fleet of horse-drawn wagons and wooden barrels. Today, its successor ConocoPhillips is worth $141.6 billion. Conoco is the quintessential independent oil exploration and production company — it doesn't run gas stations, it doesn't refine oil, it doesn't do chemicals. It does one thing: find oil and natural gas in the earth's crust, then pull it out. It is a pure "upstream company."

Not a supermajor. Not a fully integrated chain. Not refining-petrochemical hybrid.

It is "exploration and production only" — the purest oil extraction asset there is. In 2002, Conoco and Phillips Petroleum merged — forming what is now ConocoPhillips. It has no retail stations — while Shell and BP fought for the best gas station locations at intersections, Conoco took positions deep below the strata, "where no one can take your spot." In 2021 it acquired Shell's Permian Basin assets; in 2023 it bought Marathon Oil in a massive deal — continuing to double down on America's core shale oil regions. Its bet is simple: before the world fully decarbonizes, oil and natural gas remain the largest source of primary energy for humanity, and it intends to become the last great producer while the entire industry is under-investing.

Three things:

First: Don't take the retail profits of an "oil company" — only claim the upstream "geological reserves." Pure means deepest.

Second: Add to your position when others retreat from shale — the downcycle is every upstream company's inventory-clearing window.

Third: No refining and no selling gasoline to consumers — just sell crude to refineries, and let the market do the fluctuating. The purest business is the one most resilient to fluctuation.

← Previous Contents Next →
Chapter 83 Marvell-Technology

In 1995, in Berkeley, Sehat Sutardja, his wife Weili Dai, and his brother Pantas Sutardja founded Marvell Technology at their kitchen table. The first chip they designed — a fully digital hard-drive read channel — consumed one-tenth the power of the industry standard at the time. Today, Marvell is worth $141.3 billion. It moved from storage controllers into networking chips, automotive Ethernet — and, in the data center AI era, into DSPs and custom compute silicon. Marvell is hidden inside every "data transfer joint" that consumers never see.

Not application processors. Not smartphone SoCs. Not graphics chips.

It is "data processing in motion" — when a data packet moves from copper to fiber, from disk to memory, from server to server — Marvell's chips handle physical-layer signal conversion and error correction across the entire path. Sutardja himself is an analog-mixed-signal genius — he can extract decibel-level improvements in signal-to-noise ratio from the deepest physical noise at the bottom of a chip.

Three things:

First: Where the noisiest and faintest signals meet — that is where a chip is worth the most.

Second: A founding trio of three family members — one an analog genius, one managing markets and customers, one handling operations. This is the extreme collaboration of a family intelligence triangle.

Third: Every time a data interface upgrades (SATA to PCIe to optical DSP) — Marvell collects a bridge toll on that interface.

← Previous Contents Next →
Chapter 84 Interactive-Brokers---盈透证券

In 1977, Thomas Peterffy, born in Budapest, sat in a corner of the New York Commodity Exchange. He was the first person on Wall Street to replace a trader's shouted prices with a computer model. He wired an Apple II into the trading desk — while others were still using chalk and hand signals to quote prices, he was already using code to calculate the fair value of options. Today, his company Interactive Brokers is worth $141.2 billion. Peterffy was an algorithmic trader before the concept of "algorithmic trading" even existed.

Not an investment bank. Not a hedge fund. Not a payments network. Not wealth management.

It is "driving trading costs to the absolute limit" — Peterffy has believed one thing since the 1970s: human voice-pricing creates spreads and errors, and only computers can eliminate them. He first made his fortune on corn options at the Chicago Mercantile Exchange, then spent nearly a decade laying automated market-making systems and an entire electronic trading network across the globe.

Interactive Brokers does not sell investment advice — it provides "the lowest-cost electronic trading access." Its clients include professional traders, hedge funds, international investors, and quantitative teams running strategies via API. Interactive Brokers' profits do not come from its clients' losses — they come from "one of the highest trade volumes in the world, at an extremely low commission per trade."

Three things:

First: Peterffy quoted prices with an Apple II in the 1970s — that was the Paleolithic age of electronic market-making. He was his own alpha.

Second: Never bet against the client — pure agency execution only. This means it lives whether the client wins or loses.

Third: Put over a hundred global exchanges and currency markets under one screen's tradable rights — this is the highway toll station built for "capital globalization."

← Previous Contents Next →
Chapter 85 CrowdStrike

In 2011, George Kurtz and Dmitri Alperovitch founded a security company called CrowdStrike. They made an audacious bet — that all antivirus software, which relies on "virus signatures" to identify known malware patterns, was already dead in the face of cloud-native and nation-state APT attacks. They did not sell antivirus. They installed a lightweight sensor on every protected computer — sending all suspicious behavior to the cloud, where AI compared it against a global fingerprint database of attacks. Today, CrowdStrike is worth $138.8 billion.

Not antivirus software. Not a firewall. Not penetration testing. Not security consulting.

It is "transforming security from local, signature-based static defense — into cloud-based, continuous behavioral analysis dynamic defense." Kurtz himself was previously CTO at McAfee — he had seen with his own eyes how antivirus engines failed against advanced attacks. The Falcon platform rebuilt security in that space — no virus signatures, just "real-time hunting." When you read CrowdStrike's threat hunter reports — the 2016 breach of the Democratic National Committee by Russian hackers was discovered by CrowdStrike.

Three things:

First: Abandon antivirus — replace it with "cloud-based behavioral fingerprinting." This is a genetic rewrite of the traditional cybersecurity product.

Second: Replace the massive antivirus client with a single ultra-light sensor — users won't uninstall it because it barely consumes resources.

Third: Detect nation-state adversaries in real APT incidents — the company's brand becomes synonymous with the world's "digital fire alarm."

← Previous Contents Next →
Chapter 86 Honeywell---霍尼韦尔

In 1885, in Wabash, Indiana, Albert Butz invented a temperature-differential automatic regulator that allowed "cold air to rise from the basement into the living room" — later called the thermostat. Then Mark Honeywell improved the patent and founded the Honeywell Heating Company. Today, Honeywell is worth $137.5 billion. Its control sensors are inside every smart building, in the air conditioning systems of Boeing aircraft, and in the valves and instruments of every major oil refinery. It is not a consumer brand — it is the nerve endings of the industrial world.

Not consumer electronics. Not a software platform. Not a materials revolution. Not the internet.

It is "control" — temperature, pressure, speed, fluid level. Honeywell started with the thermostat, then turned "regulating every physical parameter that needs regulating" into the most foundational infrastructure of human industrial civilization. It merged with AlliedSignal — then, in a remarkable reversal, undertook massive spin-offs in 2018 and 2024, shifting its focus from a cross-industry conglomerate to one concentrated on aerospace, automation, and energy-transition materials.

Three things:

First: Evolving from a home thermostat to an aircraft cabin's air circulation system — it is the same "precise control of the physical environment" mindset, amplified infinitely.

Second: An industrial company's longevity does not depend on a single product — it depends on "reusable control theory," implanting sensors again and again in every place that requires precise regulation.

Third: Spinning off takes more courage than buying — after becoming an all-encompassing giant, Honeywell had the guts to carve away its non-core divisions.

← Previous Contents Next →
Chapter 87 Vertiv-Holdings

In 2016, Vertiv spun off from Emerson Electric and went public. But its true origin dates to the 1940s and the Liebert Corporation — founded by Ralph Liebert, the man who invented the first precision air conditioner for computer rooms. In 1965, he built the first "environmental control unit" designed for mainframe computers — not to keep people cool, but to cool IBM mainframes. Today, Vertiv is worth $135.4 billion. In every data center in the world — across thermal management, power protection, and rack infrastructure — Vertiv is the machine that keeps your cloud from turning into a puff of black smoke.

Not chips. Not servers. Not networks. Not cloud services.

It is "precisely exhausting every single watt of heat from the data center." The power density of AI data centers has exploded — a single rack has surged from 5kW to over 60kW. Liquid cooling, in-row cooling, precision power distribution — behind every GPU, a set of Vertiv cooling pipes and switches hums away.

Three things:

First: Don't build the computer — build the physical protective layer that keeps the computer at a steady 68°F.

Second: For every ten million additional AI parameters, cooling must be redesigned. This is the eternal "cooling tail chasing the explosive growth of AI compute."

Third: Seven seconds of power loss can destroy a cloud company — and Vertiv's UPS (uninterruptible power supply) handles "that seven-second physical payment."

← Previous Contents Next →
Chapter 88 Prologis---安博

In 1984, in San Francisco, Hamid Moghadam began logistics warehouse investing at AM Property. Later, he merged with Security Capital to form Prologis. Today, Prologis is worth $134 billion — it is the world's largest owner of "the big metal boxes where Amazon packages sit before they reach your doorstep." Every "next-day delivery" from your online shopping — passed through at least one Prologis warehouse along the way.

Not e-commerce. Not logistics delivery. Not trucks. Not routing algorithms.

It is "the real estate layer of logistics infrastructure." Prologis owns or manages thousands of distribution centers and logistics warehouses around the world — from the giant sorting centers next to the Port of Long Beach to the last-mile delivery stations at the edge of every major city. E-commerce growth, supply chain globalization, and the "faster delivery" arms race — everything is increasing the physical demand for logistics warehouses. And industrial land on the outskirts of major cities is growing ever scarcer — the land deeds in Prologis's hands are becoming increasingly impossible to replicate.

Three things:

First: Not in malls, not in office buildings — only in "the big concrete block where packages must pause before reaching you."

Second: Warehouses are not like residential real estate — every boost in e-commerce increases warehouse demand; every supply chain disruption (port congestion, tariffs, etc.) increases bonded warehouse demand.

Third: The scarcest non-renewable resource is not oil — it is industrial land on the outskirts of metropolitan areas.

← Previous Contents Next →
Chapter 89 S&P-Global

In 1860, Henry Varnum Poor published the first edition of History of the Railroads and Canals of the United States. He was an eccentric who standardized the financial data of American railroad companies and published it publicly — allowing investors, for the first time, to compare the financial condition of different railroads. His book inspired the logic behind all subsequent financial information services. Today, S&P Global is worth $126.6 billion — its S&P 500 index, bond ratings, and global market data form one of the operating systems of capitalism.

Not investing. Not trading. Not algorithms. Not analysis.

It is "turning the standardization of financial information into the objective medium of global capital allocation." Its bond ratings (AAA to D) determine the cost of borrowing for nations and corporations. Not advice — but "an objective risk benchmark."

Three things:

First: When you compare Railroad A and Railroad B's profit per mile on the same table — you no longer invest by gut feeling. This is "information standardization."

Second: A rating is not a moral judgment of good or bad — it is a "consistent, quantified language of default probability." The meaning of a standard lies in the word "consistent."

Third: The S&P 500 is the most cited benchmark index in the world — it doesn't pick stocks; it "defines" what a large American company is.

← Previous Contents Next →
Chapter 90 Lowe's---劳氏

In 1921, in North Wilkesboro, North Carolina, a small-town hardware store owner named Lucius Lowe turned his little shop into a one-stop building supply and general merchandise store. A hundred years later, Lowe's Companies is worth $126.3 billion. It is the second-largest home improvement retailer in the United States, behind Home Depot. Its difference from Home Depot is vanishingly small — but it strikes a gentler balance between professional contractors and the DIY crowd. Its blue sign is the first light American families see at 7 a.m. on a Saturday morning, lining up to buy pipes and garden soil.

Not Home Depot. Not full-category. Not a price war. Not a new line of business.

It is "that blue building supply store you know best, right near your home." Early on, Lucius Lowe transformed the shop from "if you need a nail, the owner has to go find it in the back" to "push your cart down the aisles and choose for yourself." The postwar baby boom and suburbanization created an endless need for home repairs across America's sprawling single-family houses — and Lowe's blue paint is locked in tight with that repair cycle.

Three things:

First: Compete for the same market as Home Depot — but always hold the differentiated position of "cleaner, quieter, more female-friendly."

Second: No building supply store wins on "lowest price" — what it sells is proximity plus a staff member who can help you read a plumbing diagram.

Third: A hundred-year retail business doesn't get beaten by e-commerce — because shipping a 12-foot roll of floor mat costs more than the profit margin.

← Previous Contents Next →
Chapter 91 Newmont---纽蒙特

In 1921, Colonel William Boyce Thompson founded the Newmont Mining Corporation in New York and immediately bought gold mines in California and Nevada. A hundred years later, Newmont is the world's largest gold mining enterprise — worth $125.1 billion. Its assets were formed, in geological time, from a surge of hydrothermal rock two hundred million years ago — and it digs that gold out of the earth's crust, melts it into bullion, and stores it in underground vaults. Gold does not rust, does not oxidize, does not dissipate — it is humanity's ultimate store of value against all monetary anxiety.

Not finance. Not technology. Not consumer goods. Not energy.

It is gold. In every wave of paper-currency inflation and every war panic, gold gets revalued. Newmont does not speculate on gold — it simply sells a bit more in the upswings, a bit less in the downswings, and allocates cash from each cycle to exploration for the next gold-bearing vein. The average time from gold discovery to production exceeds ten years — mining is not financial speculation. It is the slow engineering of "how much gold can you deliver a decade from now."

Three things:

First: Gold has no CEO — it can't be fired, can't break the law, can't issue a bad quarterly report.

Second: The extraction cycle dictates — just as everyone is buying gold ETFs, Newmont has barely decided to add one more leach pad — it profits when the next gold-price panic hits.

Third: ESG requirements are making every new mine harder and harder to permit — and Newmont holds vast quantities of already-permitted, undeveloped mining rights. That is the scarcest non-financial asset in the gold class.

← Previous Contents Next →
Chapter 92 Booking-Holdings

In 1997, Jay Walker incorporated Priceline.com in Stamford, Connecticut — a revolutionary name-your-own-price model. You didn't choose a flight — you said, "I want to go from New York to Miami, $100, any time," and airlines anonymously bid their unsold seats. This upended the airline pricing system. Later, in 2005, Priceline bought a tiny Dutch startup called Booking.com for $135 million — and at the time, every global analyst called it a bubble price. Today, Booking Holdings is worth $124.6 billion. Booking.com is the largest hotel reservation platform on earth, and Priceline's original star has become Booking Holdings.

Not an OTA innovation. Not metasearch. Not travel reviews. Not flight-plus-hotel bundling.

It is that one deal — "bought a Dutch company called Booking.com that looked tiny." A Dutchman named Steyn Nordhoek founded Booking — he turned online hotel booking into "pay at the hotel, cancel any time, no prepayment" — with the platform fronting the payment risk to the hotel. In the 2000s, this was extremely risky because it meant a large number of no-show charges. But when consumers no longer had to "worry about cancellation penalties" — bookings rose far faster than the penalty costs. This user-experience shift was the singularity point: Booking.com devoured every traditional hotel reservation channel.

Three things:

First: Use "cancel anytime" to zero out the consumer's booking anxiety — the increase in order volume more than offset the no-show rate.

Second: In 2005, pay an "outrageous price" for a no-name Dutch company — while everyone else's eyes were glued to Expedia and TripAdvisor.

Third: It is not a "travel company" — it is "a two-sided marketplace for precisely matching accommodation prices."

← Previous Contents Next →
Chapter 93 Starbucks

In 1983, a kitchen-supply salesman from Brooklyn flew to Milan for a housewares trade show. He walked into his first Italian coffee bar beside the Milan Cathedral — stood at the counter and watched for three minutes: a barista turned espresso and steamed milk into a brown swirl in under sixty seconds, then slid the cup across to a man in a suit. The man drank it standing, gave the barista a nod, done in thirty seconds. That salesman stood in front of his little espresso cup, and something detonated inside his head. Not because the coffee tasted good. Because he saw it — coffee was not a beverage. Coffee was the only conversation each morning between one lonely person and another.

Not discovering coffee. Not improving roasting. Not inventing espresso.

It was Howard Schultz. He returned to Seattle and told Starbucks's three founders: we can't just sell coffee beans anymore. We have to sell "the Italian coffee experience." The founders said no. He quit. He opened his own Italian coffee bar, called Il Giornale. A year later, he bought his old employer and renamed it — Starbucks.

Today, Starbucks is worth $121.1 billion, with 38,000 stores worldwide.

Let me tell his story first. After I finish, we'll talk about why the business concept of "the third place" — is not a room, but a stretch of warm alone-time.

Howard Schultz grew up in the projects of Brooklyn. His father was a blue-collar worker who did odd jobs — driving trucks, delivering diapers — and when Howard was seven, his father broke his leg, got fired, no insurance, no workers' comp. Every morning as a child, he woke to the smell of strong disinfectant mixed with leftover food in the hallway — the fixed scent of public housing. Their family had no phone. A neighbor let them use theirs, and when a call came, the neighbor would shout down the corridor: "Schultz — phone —"

Schultz won a football scholarship to Northern Michigan University. After graduation, he went to Xerox as a salesman. Then he jumped to a Swedish kitchenware company, selling drip coffee makers. He noticed a small shop in Seattle — selling only roasted coffee beans — had ordered their company's top-of-the-line drip brewer, in quantities far larger than any department store. He decided to fly to Seattle and see.

It was the first Starbucks store. Founded in 1971 by Jerry Baldwin, Gordon Bowker, and Zev Siegl at Pike Place Market. They sold only freshly roasted coffee beans and grinding equipment — no cups of coffee, no beverages. The shop was filled with the smell of burnt caramel from the roaster and the sound of the grinder. Schultz stood in that shop for an entire afternoon, listening to the three founders discuss with customers the subtle differences between Sumatran and Guatemalan beans — the customers spoke as if they were discussing wine.

In 1982, he joined Starbucks to run marketing. In 1983, he went to Milan — and that cup of espresso pushed across the counter to him in the early morning changed the entire course of his life from that moment on. He later wrote in his memoir: "That cup of coffee — it was handed from one person's hand to another. Not dispensed automatically by a machine. A person. A person who was simultaneously making every drink for seven or eight people gathered around the counter, while chatting with each and every one of them."

Back in Seattle, he argued repeatedly with the founders: we should serve freshly brewed coffee and espresso drinks in every store. The founders refused — "We are bean sellers. We are going to be the greatest importer of Colombian coffee. We do not run restaurants."

In 1985, Schultz left Starbucks and opened his own Il Giornale in Seattle. While raising money, he was rejected by banks and investors over 240 times. In the end, it was an Italian espresso machine manufacturer — "You're going to do Italian coffee — I believe in you."

Il Giornale opened its doors. By 7 a.m., the line stretched all the way into the afternoon. In 1987, the three Starbucks founders decided to sell the company. Schultz scraped together $3.8 million, bought Starbucks — and merged Il Giornale into it, unifying everything under the name Starbucks.

On the first day after the merger, he hung a photograph of an Italian coffee bar on the back wall of every store — and said to the staff a line that every Starbucks trainer has quoted ever since: "We are not in the coffee business serving people. We are in the people business — serving coffee."

Over the next 20 years, Starbucks became one of the most prominent street-level sights in America. Schultz's core logic was a single idea: "the third place" — every person needs two places: home (the first place), the office (the second place), and "a safe transitional zone between the two, belonging to neither." Here you can sit alone for an entire afternoon, no one will rush you, and the person behind the counter might even know your name.

This is not a function of coffee — this is a sociological invention.

In 2008, Starbucks was hit by a triple blow: over-expansion, a cluttered menu, and the financial crisis. Schultz, already retired for eight years, returned as CEO. He did something that sent the whole company into a panic: at 5:30 p.m. on the same day — he shut down all 7,100 Starbucks stores in America — to retrain every barista on the standards of pulling espresso. It was Starbucks operating on itself — admitting "our coffee has become mediocre."

After the retraining, the menu was cut, milk pitchers returned to made-to-order, the hot dog racks and the burnt-cheese smell of breakfast sandwiches were removed. And Starbucks came back to life. Then it expanded into China, into India, into every white-collar gathering place hungry for an hour and a half of uninterrupted Wi-Fi.

Stop here for a moment.

Have you ever given your customers — beyond your core product or service — a "third place"? Not a physical space. A "psychological transition zone where one can safely be alone, or exchange a few words with a stranger." If your industry vanished tomorrow — would your customers miss not just your product, but "that stretch of time they spent with you"?

Three things.

First: What you're selling is not coffee — it's "being recognized" on the other side of the counter each morning.

Schultz has always insisted that Starbucks is not called "fast food." Because a barista should remember a regular's name and drink — this is manual, slow, anti-scale. Can your staff naturally address your customers by their last names?

Second: Shut down every store for a day — to repair the standard.

The 2008 closure for retraining was Schultz's most criticized decision — tens of millions in revenue, thrown away in a single afternoon. But he said, "If you don't patch the leak in trust — what drains away every day after is not just money, it's every habit." Would you be willing to stop one day's revenue — to fix a systemic quality problem?

Third: The "dignity mechanism" of a project kid.

In his memoir, Schultz has said over and over that he gave all Starbucks employees — including part-timers — health insurance and stock options because of the humiliation he saw through his seven-year-old eyes when his father broke his leg and was fired. He would never, in his life, become that kind of boss. Starbucks was one of the first chain companies in America to provide health insurance to all part-time employees — not because of the law, but because of a seven-year-old boy's memory of watching his father lose his job.

In your company's policies — is there one that grew out of the most painful stone of your childhood?

At the very end of his memoir, Howard Schultz wrote a final line: "Coffee drips through, drop by drop. But connection with a person — that is handed from one person's hand to another. A machine can never do that."

← Previous Contents Next →
Chapter 94 Lockheed-Martin

In 1926, Los Angeles. A pair of brothers started building airplanes in a rented garage. The older brother was Allan Lockheed; the younger, Malcolm. Neither had an engineering degree. Their first plane was cobbled together in a borrowed soundstage on a Hollywood film lot — the fuselage was glued plywood. Then they broke Charles Lindbergh's long-distance flight record. Today, Lockheed Martin is worth $119 billion. It is the manufacturer of the world's most advanced stealth fighters, missile defense systems, and space reconnaissance platforms.

Not the military-industrial complex. Not government relations. Not political lobbying. Not war-driven.

It was the Lockheed brothers — and later, Clarence "Kelly" Johnson. A Swedish-American engineer born in a Michigan mining town, who once worked as a die-maker's apprentice at the Ford Motor Company. In 1943, next to the Burbank plant, he rented a windowless warehouse and used curtains and wooden partitions to wall off a sealed room. Inside, he placed 23 engineers and 30 mechanics — all handpicked — and bet everything on a jet fighter that had no name yet. He gave them a line that every Silicon Valley company would later borrow — "Be quick. Be quiet. Be on time."

That warehouse is now called the Skunk Works. The SR-71 Blackbird it produced remains the fastest air-breathing aircraft in human history. The F-117 stealth fighter was the first combat aircraft invisible to radar. The F-22 and F-35 are the pillars of today's air superiority.

The core of this story is not war — it is "putting the fastest, quietest minds into a small dark room with the fewest layers of management, and then giving them an impossible limit."

Kelly Johnson, by age 14, could use the family bellows and scrap iron to forge small knives and tin airplane models. He swept the floors of the University of Michigan's wind tunnel lab — scrubbing test tubes and mopping floors in exchange for the right to sleep beside the wind tunnel. Later, he joined Lockheed. His first major recorded contribution — he used pliers and thin steel plate to manually rework the rudder design of the Lockheed Model 10 Electra, turning a commercial airliner that was about to crash into the fastest twin-engine aircraft of its time. Then came World War II.

In 1943, British intelligence issued an urgent request — they needed a reconnaissance-fighter hybrid platform that could fly over Berlin without being caught by German anti-aircraft missiles and fighters. The ceiling and speed of every existing fighter fell short. Kelly selected the P-80 Shooting Star as the base but essentially redesigned every core element — and gave Britain and the U.S. Army Air Forces a pledge: deliver the prototype within 180 days. Actual delivery: 143 days. The production run was small, but it cemented the Skunk Works' working philosophy:

Kelly laid down his Fourteen Rules — the first being: "The Skunk Works program manager must have nearly complete control over all aspects of the program. He reports directly to the president or senior management." And also — "The number of reports that must be written must be restricted, but important work must be documented in extreme detail." And — "Only the smallest possible number of outside inspection teams shall be allowed into internal processes."

At the time, this was an "anti-company within the company" — seal the elite inside a dark room, produce physical hardware on an impossible timeline, then let them fly the test flights themselves. He himself climbed into the test aircraft more than once.

The Skunk Works' output reshaped every air operation over the following decades: before the U-2 spy plane was shot down over Soviet airspace, Soviet altitude radar could not even see it. The SR-71 Blackbird was never shot down — its primary defense was flying faster than missiles. Once locked on, it did not fire — it accelerated to Mach 3 and flew straight out of the threat envelope.

Then, in the late 1970s, the F-117 stunned the world with its multi-faceted angular shape — it barely reflected any radar waves. During the 1991 Gulf War, the F-117 flew into the skies over Baghdad, struck its targets amidst the densest concentration of anti-aircraft fire on the planet, and took zero losses.

In 1995, Lockheed merged with Martin Marietta. Then came the F-35 Joint Strike Fighter — over five thousand units in production, making it the most expensive weapons system in history, and simultaneously the most widely deployed fifth-generation stealth aircraft in the world today. Debates and revisions around its tactical networking and data-fusion capabilities continue to this day.

Stop here for a moment.

Have you ever given your most critical project — a separate, physically isolated space — with no fixed closing time, file permissions walled off from headquarters, keys held by only three handpicked people?

The core of Kelly Johnson's Fourteen Rules was not "cost reduction." It was "kicking everything that even looks like 'management' out of the room where the product is born." Where is your "Skunk Works" — the thing that can do impossible missions?

Three things.

First: Kick every layer of management that can slow things down out of the room.

Kelly believed management is a form of friction — it turns fast into medium, unease into comfortable delay. So he had the Skunk Works report directly to the highest decision-maker — one single level. On your urgent project — how many layers of "middle management" sit between you and the decision-maker, any one of which can pass down the word "no"?

Second: Set yourself a deadline that looks impossible — then announce it publicly.

143 days. Kelly didn't calculate it. He simply said "180 days" — and then the engineers, in terror, began reverse-calculating the fabrication window for every single component. Do the deadlines you give your team fall inside or outside what they think is feasible?

Third: Stealth is not magic — it comes from deconstructing the minimum physical reflectivity.

Every facet of the F-117 is a physical calculation of radar-wave backscatter angle. It's not "stealth magic" — it's pure physics of angles and materials. Can you find an opportunity in your industry to "use basic physics to bypass others' complex strategies" — don't play their game, invent your own geometry?

Kelly Johnson passed away in 1990. His desk still sits inside the original Skunk Works building, topped with a Plexiglas plaque inscribed in handwriting: "Be quick. Be quiet. Be on time."

← Previous Contents Next →
Chapter 95 CVS-Health

In 1963, in Lowell, Massachusetts, brothers Stanley and Sidney Goldstein and their partner opened a small health and beauty shop called Consumer Value Store — CVS for short. The first store sold toothpaste, shampoo, and bandages. Then, step by step, it swallowed the pharmacy next door, the clinic, and the insurance business. Today, CVS Health is worth $118.9 billion — from chain pharmacy to pharmacy benefit management (PBM) to in-store MinuteClinics, it has become the "pre-hospital" entry point that covers tens of millions of Americans' daily medical and prescription needs.

Not pharmaceuticals. Not insurance. Not hospitals. Not tech innovation.

It is "that drugstore on your corner — which is now, at the same time, your vaccination site, your chronic-condition check-up stop, and the entity that negotiates your drug benefits on your behalf." CVS first bought Caremark (PBM), then bought Aetna (health insurance), placing the payer and the drug dispenser under one roof. You are an Aetna insurance member — the price of your prescription at a CVS pharmacy automatically comes out lower than anywhere else. This is a closed loop.

Three things:

First: From "selling toothpaste" to "managing the closed loop of every prescription pick-up and insurance claim you make" — a triple jump, all completed under the same physical roof.

Second: The expansion of MinuteClinics inside stores — stripping "seeing someone for a minor illness" out of the hospital appointment queue and placing it into a retail setting.

Third: Vertical integration — it is pharmacy + PBM + insurance, capable of negotiating with itself inside its own system.

← Previous Contents Next →
Chapter 96 Danaher---丹纳赫

In 1984, brothers Steven Rales and Mitchell Rales founded a small investment company in Massachusetts. Then they began an unending spree of acquisitions in life sciences and industrial technology — using an extremely rigorous efficiency management methodology called the "Danaher Business System." From water quality analyzers to mass spectrometers to single-use bioreactors for biopharma — it buys the world's "scientific research tools and diagnostic hardware," feeds them into a lean manufacturing mold, and squeezes out exceptionally high profit margins. Today, Danaher is worth $118.6 billion.

Not scientific breakthroughs. Not pharmaceuticals. Not clinical care. Not original technology.

It is "the lean philosophy of manufacturing, applied to scientific instruments." Danaher is not one company — it is a methodology attached to a never-ending acquisition machine. Its "Danaher Business System" is a comprehensive toolkit for operational improvement, built on the Toyota Production System and lean manufacturing — analyzing every operational node of every acquired company, compressing waste, and accelerating cash conversion. Then it uses the money saved to buy the next one.

Three things:

First: Don't invent the science — give scientists the tools, then make those tools as stable and drift-free as possible.

Second: DBS is not a corporate slogan — it is a mandatory course in every employee's first week and the foundational architecture of their performance evaluation.

Third: Perpetual acquisition — when your operating system can systematically increase return on net assets, your acquisitions themselves become a compoundable financial instrument.

← Previous Contents Next →
Chapter 97 Altria---奥驰亚

In 1847, on Bond Street in London, Philip Morris opened a small shop selling hand-rolled Turkish cigarettes. Over a century later, his brand Marlboro became the most valuable tobacco brand in the world — bound to the image of "the cowboy." Today, its parent company, Altria Group, is worth $116.6 billion. At the same time, the tobacco company's "transformation" has become the world's most contentious and seductive business narrative — it legally acknowledges that smoking kills and pays out billions in settlements, while using a portion of its profits to buy Juul, buy the smoke-free nicotine pouch On!, and buy heat-not-burn technology. It sells nicotine, while saying itself that the future should involve no more sales of combustible cigarettes.

Not health. Not technology. Not social justice. Not medicine.

It is nicotine. An alkaloid that simultaneously addicts hundreds of millions of people and is treated by public health as a slow-moving killer. Altria's future business model is essentially a paradox — it publicly declares it is moving "toward a smoke-free future," but its core profits still come from the tar and thousands of chemicals produced by the combustion of every pack of Marlboros.

Three things:

First: When your core product is proven to cause harm — do you choose transformation or defense? Altria is "doing two opposing things at the same time."

Second: Nicotine pouches and heated tobacco are considered "less harmful than combustible cigarettes" — does this count as "risk shifting" in public health?

Third: The power of the Marlboro cowboy's visual legacy — long after tobacco advertising was banned by law, those red-and-white racing sponsorships and old advertising archives can still automatically ignite a nicotine craving inside the human brain.

← Previous Contents Next →
Chapter 98 Progressive---前进保险

In 1937, Jack Green and Joseph Lewis founded Progressive Insurance in Cleveland. They were the first to offer "installment premiums" — allowing people, most of whom were paid weekly at the time, to pay for auto insurance weekly or monthly, instead of in one lump sum for the entire year. Today, Progressive is one of the largest personal auto insurers in the United States, worth $116.1 billion.

Not life insurance. Not health insurance. Not property insurance. Not reinsurance.

It is "drilling auto insurance data down to the finest granularity of actuarial precision." Progressive was a pioneer of the UBI (usage-based insurance) onboard diagnostic device — it feeds a driver's real driving data (number of hard brakes, average speed, time-of-day driving patterns) into the pricing formula, abandoning crude statistical categories. Safe drivers thus get low rates, while high-risk drivers get premiums that more accurately reflect their risk.

Three things:

First: Insurance is not "selling peace of mind" — it is "pricing every individual's risk so fairly that the person themselves cannot refute it."

Second: Weekly installment payments — letting the poor get on the road legally — this is the low-barrier realization of actuarial justice.

Third: UBI transforms insurance from a "claims game" into a "driving behavior feedback loop" — the second you start changing your driving habits to lower your rate, you have already become its long-term customer.

← Previous Contents Next →
Chapter 99 Bristol-Myers-Squibb---百时美施贵宝

In 1887, in Clinton, New York, two young men — William McLaren Bristol and John Ripley Myers — founded the predecessor of Bristol-Myers Squibb in the back room of a rented pharmacy. They didn't invent drugs — they did something stranger: they took effective drugs already on the market, made them into purer crystals, and pressed them into gelatin capsules that were easier for patients to swallow. Today, BMS is worth $115.4 billion, and its immuno-oncology drugs Opdivo and Yervoy have rewritten cancer treatment.

Not biotech breakthroughs. Not gene editing. Not targeted therapies.

It is "taking drugs others invented but that were hard to take — and making them into easier-to-swallow dosage forms." Sulfa drugs. Antibiotics. Blood pressure medications. Then BMS used the cash accumulated from its own formulation capabilities to acquire a series of blockbuster biologics and oncology pipelines from the 1990s to the 2020s. After merging with Squibb in 1989, it became BMS — and then turned immuno-oncology into the third pillar of human cancer treatment (after surgery and chemotherapy).

Their "immune checkpoint inhibitors" do not directly kill cancer cells — they strip away the "fake police badge" (PD-1/PD-L1 blockade) that cancer cells use against immune T-cells — enabling the patient's own immune system to recognize and devour the tumor.

Three things:

First: Started by "improving the dosage forms of other people's drugs" — turned "ease of administration," something no one else cared about, into a hundred-year cash machine.

Second: Cancer is not about killing it with poison — it's about "releasing the immune brakes" so your own body goes and bites it.

Third: PD-1 inhibitors proved a principle — sometimes the greatest medical breakthrough is not "inventing an attack," but "turning off an inhibitory signal that was preventing your own attack."

← Previous Contents Next →
Chapter 100 Vertex-Pharmaceuticals---福泰制药

In 1989, Cambridge. Joshua Boger founded Vertex Pharmaceuticals. He was 32 years old, a madman for something called "rational drug design" — not relying on mass screening of compound libraries and hoping for luck, but on computational molecular simulation. For nearly the entire first decade, every lead compound failed in animal studies or early clinical trials. Then, in 1999, he threw Vertex's fate onto an unprecedented bet: to try to cure, using an all-oral "synergistic combination therapy," a chronic disease that had been considered "manageable but never eradicable for the last hundred years" — hepatitis C. He lost (Gilead was faster). But it was precisely the structural biology and enzyme inhibitor platform accumulated in that failure that later produced the first corrective drug in human history targeting the disease-causing gene of cystic fibrosis — Kalydeco, Orkambi, and then Trikafta — turning a genetic death sentence with an average age of death under 40 into a chronically maintainable state compatible with a normal life. Today, Vertex is worth $114.6 billion.

Not chemical drugs. Not biologic antibodies. Not gene editing. Not immunotherapy.

It is "for every unique CFTR mutation subtype — design a small-molecule corrector, place it into the misfolded protein groove caused by that mutation — and then it works normally." Every patient must be tested for the mutation type they carry to determine which corrector series will be effective. This is not a broad-spectrum treatment of one drug for ten thousand people — it is one mutation, one drug.

Three things:

First: From the hepatitis C failure of "losing to Gilead" — recover and repurpose the accumulated technical expertise in structural biology and protein folding, pivoting to cystic fibrosis.

Second: Not treating symptoms — a few pills a day can "repair the fundamental protein folding defect caused by the genetic mutation."

Third: The market for rare diseases is too small? When you can turn the hundreds of thousands of dollars in annual drug costs for families whose children "only have 10 to 15 years of life left" into a nursing-care economic structure — the model is viable.

← Previous Contents Next →
Chapter 101 Capital-One

In 1988, Richard Fairbank and Nigel Morris founded Capital One in Virginia. They were not traditional bankers — Fairbank was a Stanford MBA obsessed with applying statistics and experimental methods to the credit card business. He ran thousands of simultaneous A/B tests every year — interest rates, colors, wording, credit limits — using statistical significance to optimize every single envelope he sent out. Today, Capital One is one of the largest consumer banks in America.

Not a traditional bank. Not an investment bank. Not a mortgage lender. Not wealth management.
It was "turning the credit card into a vehicle for statistical experimentation." Its core IP was not a banking charter — it was the machine learning models and experimental database, built over decades, on the repayment behavior of American consumers. Capital One is, in truth, "an AI lab that does credit modeling, which happens to issue cards."

Three things:
First: It didn't rely on "a lender's intuition" — it relied on "running tens of thousands of simultaneous A/B tests to statistically determine which customers would repay."
Second: A bank's most valuable asset isn't its vault — it's "who can most accurately predict who won't default."
Third: Capital One's founding logic is everywhere in fintech today — it was the earliest fintech company, 25 years before the word existed.

← Previous Contents Next →
Chapter 102 Quanta-Services

1997, Houston. Quanta Services got its start in oilfield power and telecommunications cable engineering. Then renewable energy and electrification exploded. Today it is the largest specialty engineering and construction contractor for power transmission and renewable energy in America — wherever someone needs to build a new solar farm, lay a high-voltage underground cable, or install EV charging stations and step-up transformers — Quanta is the actual builder on the ground with the excavators and cable trucks. Its market cap exceeds one hundred billion dollars.

Not a power utility. Not an equipment manufacturer. Not a renewable energy developer.
It was "you can actually get electricity from the solar panels in the desert to the outlet in the city — every single utility pole and every meter of underground cable required in between." Quanta's workers climb high-voltage poles in mud-caked work boots — while the entire world is turning every form of energy into electricity.

Three things:
First: America's aging grid plus the demand for renewable energy transmission — grid modernization is a real-world engineering project spanning hundreds of thousands of miles. This isn't cyber. It's real copper and steel.
Second: The physical bottleneck in the electrification transition isn't blueprints — it's "enough cable splicers and underground drilling rigs."
Third: No matter who generates the power or who sells it — the cables between the power plant and the end user have to be laid by someone.

← Previous Contents Next →
Chapter 103 Stryker---史赛克

In 1941, Dr. Homer Stryker — an orthopedic surgeon in Michigan — grew so frustrated with the operating room equipment he saw that he welded a new type of fracture fixation bed in his garage, using scrap stainless steel and old bed frames. He went on to design orthopedic saws, hip implants, and mobile hospital beds. Today Stryker's market cap exceeds one hundred billion dollars — and in virtually any orthopedic or spine surgery performed anywhere on the planet, a Stryker product is almost certainly present.

Not pharmaceuticals. Not insurance. Not diagnostics. Not imaging.
It was "the saw, the drill, the fixation plate, and the joint in the surgeon's hands." Dr. Stryker's own conviction — "doctors should design the tools" — permeates the entire company. Most of Stryker's products did not come from a design lab — they came from "a surgeon saying, 'If I had a curette that could bend like this — I could do this knee procedure through one fewer incision.'"

Three things:
First: The best designer at an instrument company — is the surgeon standing at the operating table.
Second: In a single surgery, Stryker products might be present in the anesthesia equipment, the electrocautery, the fixation plates, and the postoperative irrigation — not through one big sale, but through repeated consumption across the same surgical drape.
Third: It doesn't make "high-tech." It makes tools that "get used a thousand times a day and never fail." In the operating room, reliability matters ten times more than innovation.

← Previous Contents Next →
Chapter 104 Intuit---财捷

In 1983, Scott Cook sat at the dining table of his Palo Alto apartment. His wife had covered the table with checks and bills — complaining about how agonizing household bookkeeping was. Cook was a Harvard MBA who had just quit Procter & Gamble. He picked up the phone and called banks and software companies — asking, "Why isn't anyone doing personal finance software on the personal computer?" Then he wrote Quicken — a home checkbook program rendered in black-and-white pixels on MS-DOS. Today Intuit's market cap is roughly 130 billion dollars — TurboTax and QuickBooks cover the core financial behaviors of small businesses and individuals across America.

Not a bank. Not an accounting firm. Not wealth management. Not payments.
It was "turning the United States tax code — one of the most complex documents ever written by human hands — into something where you answer simple questions on your phone screen and it automatically calculates your refund." And turning a small business's ledgers and invoices into a single click.

Three things:
First: It didn't invent the tax code — it just made it so ordinary people could get a legal refund without actually having to read the tax code.
Second: Small business accounting shouldn't require an accounting degree — QuickBooks buried the complexity of double-entry bookkeeping behind a button.
Third: Every year before April 15th, when tens of millions of Americans open TurboTax, Intuit becomes "the default infrastructure of tax-filing behavior."

← Previous Contents Next →
Chapter 105 Parker-Hannifin---派克汉尼汾

In 1917, Arthur Parker built the first reliable pneumatic brake fitting for trucks in a small machine shop in Cleveland. A hundred years later, Parker Hannifin is the most critical invisible giant in motion and control technology worldwide — from hydraulic systems for commercial aircraft to pneumatic valves for factory robots, to seals inside renewable energy equipment. Its market cap exceeds one hundred billion dollars.

Not consumer goods. Not electronics. Not pharmaceuticals. Not the internet.
It was "making fluids and motion precisely controllable in every kind of extreme environment." Hydraulics, pneumatics, filtration, sealing — Parker's products work in the unseen spaces inside machines, but they are what lift an excavator's arm, retract a passenger jet's landing gear, raise and lower a surgical table — they are "metal muscles."

Three things:
First: It doesn't build whole machines — it builds the hydraulic muscles and control nerves that "let other machines move their own bodies."
Second: Diversified end markets hedge the cycle — when aerospace declines, industrial and alternative energy may be rising.
Third: A rubber O-ring is easy to overlook — but when one fails on a Boeing control surface, it's a catastrophe. The most inconspicuous parts, bearing the highest-stakes physical responsibility.

← Previous Contents Next →
Chapter 106 Howmet-Aerospace

1888, Pittsburgh. A lineage tracing back to one of the founders of Alcoa, the aluminum giant. Howmet's predecessor was Alcoa's precision casting division. It was later spun off as Howmet Aerospace. Today it is the global leader in precision castings for aircraft engines and gas turbines. Inside every turbine blade on every Boeing and Airbus engine — those single-crystal nickel-based superalloys that hold their shape in gases hotter than their own melting point — the ceramic-mold casting was done by Howmet.

Not the whole engine. Not the design. Not the avionics. Not the assembly.
It was "making a piece of metal stay solid and keep spinning inside a flame that exceeds its own melting point." Aircraft engine turbine blades are the only mass-produced human artifact that operates in "an impossible temperature range." Howmet uses directional solidification and single-crystal casting to make every blade a single continuous crystal lattice.

Three things:
First: The frontier in materials and manufacturing processes isn't "better" — it's "functioning at the very last degree of temperature that the laws of physics permit."
Second: The fastest way to improve fuel efficiency with each generation of engine — is to raise the turbine inlet temperature. Higher temperatures demand stronger blades.
Third: Only a tiny handful of companies on Earth can produce single-crystal turbine blades — this isn't oligopoly by design, it's a physics barrier so high that no one else can jump in.

← Previous Contents Next →
Chapter 107 Equinix

In 1998, Al Avery founded Equinix in Redwood City, California. His idea was strange: not selling servers, not selling network cables — selling "the place where networks shake hands with one another." Equinix's data centers are the cross-connect points for all the cloud service providers, telecom carriers, CDNs, and banking networks. Today Equinix's market cap is near one hundred billion dollars — a WeChat message you send might not pass through Equinix, but an electronic trade from New York to London almost certainly does.

Not the cloud. Not telecom. Not fiber. Not content.
It was "interconnection" — letting AWS, Google Cloud, and Microsoft Azure peer directly inside a single building, pushing latency in every direction down to microseconds. Its neutral colocation data centers are "the transportation hubs of the internet."

Three things:
First: It doesn't sell cloud — it sells "that neutral corner where all the clouds shake hands with one another at high speed."
Second: Network effects compound — once most of the important network nodes are already inside an Equinix data center, newcomers have no choice but to enter too.
Third: Financial trading platforms, DNS root servers, CDN edges — physically, they all crowd together inside the same Equinix facilities.

← Previous Contents Next →
Chapter 108 Constellation-Energy

In 2022, Constellation was spun off from Exelon and listed publicly. But its lineage traces back to Baltimore Gas and Electric — founded in 1816, America's first gas lighting company. Today, Constellation is the largest producer of carbon-free energy in the United States — it owns dozens of nuclear power plants. In an era when AI is driving an explosive surge in electricity demand, the property of nuclear energy — "24/7 carbon-free, runs whether the wind blows or the sun shines" — has suddenly become profoundly strategic. Its market cap exceeds one hundred billion dollars.

Not technology. Not solar panels. Not batteries. Not carbon capture.
It was nuclear fission — splitting a uranium atom's nucleus in two, and letting the steam drive a turbine. Constellation's nuclear generating units are relics of a previous generation's infrastructure — but they have now become "the continuous baseload backbone power that AI data centers crave most."

Three things:
First: When the wind doesn't blow and the sun doesn't shine — the nuclear reactor is still outputting power.
Second: Existing nuclear units are already-built "sunk-cost assets" — restarting them and extending their licenses is far cheaper than building new ones.
Third: AI training and inference have caused electricity demand to surge — and nuclear is the only existing power source that can simultaneously satisfy "continuous + carbon-free + at massive scale."

← Previous Contents Next →
Chapter 109 Southern-Company

In 1912, James Mitchell founded Alabama Power in Birmingham, Alabama. This was the predecessor of what would become Southern Company. Today, Southern Company is one of America's leading electric utilities, and at Plant Vogtle it completed the country's first new nuclear reactors in decades — Units 3 and 4. From groundbreaking to commercial operation, Vogtle spanned well over a decade and endured countless cost overruns — it was one of the hardest engineering projects in the history of American nuclear construction. But in the end, it is generating power.

Not a tech company. Not oil and gas. Not a new-energy disruptor.
It was "reliable power supply" — keeping the air conditioning running through an Atlanta midsummer, keeping the auto plants from stopping on the line in Mississippi. Southern Company grew slowly but steadily in conservative Southern states under the regulated utility model. Vogtle Units 3 and 4 — no matter how much they overran — are now providing clean baseload power every single day. Outside of China and Russia, building new nuclear reactors in the United States and Europe has become extraordinarily difficult — Southern Company is one of the very few institutions that actually pulled it off.

Three things:
First: On a project with the most complex regulation, the most litigation, and the costs most prone to spiraling out of control — they still finished it.
Second: A utility is not a growth revolution — it is the foundational stabilizer that "society cannot afford to stop."
Third: While every AI company is anxiously asking "where will the power come from" — Southern Company already has the Vogtle dual units online and delivering.

← Previous Contents Next →
Chapter 110 CME-Group---芝商所

1848, Chicago. Eighty-three grain merchants gathered in a wooden building by the river and founded the Chicago Board of Trade (CBOT). They did one thing — they turned wheat, corn, and oats into standardized "futures contracts," rather than haggling over every batch at a different quality. From that moment on, commodities became financial instruments that could be traded, hedged, and priced. Later, CBOT merged with the Chicago Mercantile Exchange (CME) to become CME Group — the largest derivatives exchange in the world, with a market cap near one hundred billion dollars. Treasury futures, S&P 500 futures, EUR/USD futures, crude oil futures — all priced here.

Not a bank. Not algorithmic trading. Not a broker-dealer. Not a brokerage.
It was "standardized contracts and central counterparty clearing" — taking your anxiety about interest rates, exchange rates, and oil prices and turning it into a standard contract you can trade on margin. CME's clearing house handles trillions of dollars in margin flows and variation margin every single day — it isn't betting on the market; it is "the neutral bookkeeper for all bets."

Three things:
First: It doesn't predict the market — it gives everyone who wants to predict a single standardized venue to place their bets, hedge, and concede defeat.
Second: Central counterparty clearing — when one party defaults, the clearing house steps in with its own capital. This concentrates systemic risk inside a single, highly regulated fortress.
Third: The standardized contract invented by 83 wheat farmers a century and a half ago — the logic remains unchanged to this day, only the underlying has shifted from wheat to Treasury bonds and Bitcoin futures.

← Previous Contents Next →
Chapter 111 Cadence-Design-Systems

1988, San Jose. A group of EDA engineers from Berkeley and Bell Labs founded Cadence. What they made were "design automation tools for chip designers" — before you draw a single transistor line, Cadence's software has already simulated tens of thousands of physical effects in simulation. Today Cadence Design Systems has a market cap exceeding one hundred billion dollars. For the physical implementation of every AI chip — you must choose between Synopsys and Cadence. The two of them monopolize the world of chip design software.

Not chips. Not lithography. Not wafer fabs. Not IP cores.
It was "before the zeros and ones ever become silicon — first simulating every line width, every parasitic capacitance, and every signal delay in their entirety inside the digital world." Without clearing Cadence's timing closure and physical verification — TSMC and Samsung will not even accept your wafer order.

Three things:
First: EDA is the "first propulsion" of the entire semiconductor industry — it manufactures nothing, but without it nothing can be manufactured.
Second: The further chips push below 3 nanometers — the more complex the physical modeling becomes, and the more irreplaceable EDA grows.
Third: The Cadence-Synopsys duopoly — competition rarely comes from a third player; the iterative momentum comes from the insatiable demand AI chips place on EDA tools.

← Previous Contents Next →
Chapter 112 Adobe

In 1982, two engineers from Xerox PARC — John Warnock and Chuck Geschke — quit and founded Adobe in a garage. They took with them an invention Xerox had no interest in commercializing: PostScript, a page description language that described fonts using mathematical curves. Then Steve Jobs of Apple asked them to build the controller for the first Macintosh LaserWriter. Today, Adobe's PDF, Photoshop, Premiere, and After Effects — are the first menu items opened every single day by every designer, photographer, and video editor on the planet. Its market cap exceeds one hundred billion dollars.

Not hardware. Not an operating system. Not a social network. Not games.
It was "standardizing the entire digital workflow of human visual creation — from the first pixel to the printed page." PostScript turned fonts from "dot matrices" into "Bezier curves." Photoshop turned the darkroom into layers. PDF locked paper documents into a digital format that couldn't be tampered with. And now, Firefly is embedding generative imagery into the underlying engine.

Three things:
First: It is forever defining "the standard format for visual content" — PostScript, PDF, PSD. The power to define a standard > a software feature list.
Second: Xerox PARC invented PostScript, but Xerox thought "nobody needs to print high-quality documents from a computer." Warnock and Geschke said, "Everyone needs this."
Third: The SaaS transformation — from selling perpetual-license boxed Creative Suite to Creative Cloud subscriptions, turning one-time revenue into a perpetual annuity.

← Previous Contents Next →
Chapter 113 Synopsys

In 1986, Aart de Geus founded Synopsys in North Carolina's Research Triangle Park. He had previously worked on chip design automation at General Electric — GE did not consider EDA to be core. He took the logic synthesis tools and made them the cornerstone of Synopsys. Today, Synopsys has a market cap exceeding one hundred billion dollars, and in 2024 it announced the acquisition of Ansys for 35 billion dollars — bringing chip-level simulation and system-level multi-physics simulation together under a single roof. If Cadence is the left brain of EDA — Synopsys is the right brain.

Not chips. Not manufacturing. Not verification. Not testing.
It was "automating the entire chip design flow from RTL code to physical place-and-route." Synopsys's logic synthesis and timing signoff are the mandatory path for every digital chip. Then, through acquisitions, it entered software security and automotive software — extending "chip design verification" all the way to "verification of the software for an entire autonomous vehicle."

Three things:
First: Chip design has become too complex — a single 5-nanometer chip contains tens of billions of transistors. Without logic synthesis and physical synthesis, it would simply be impossible to finish drawing it. Synopsys's tools are the only pen that can.
Second: AI-driven chip design — Synopsys DSO.ai — letting reinforcement learning explore the physical design space. AI is designing AI chips.
Third: The Ansys acquisition — unifying chip-level and system-level simulation. Chips will no longer be designed in isolation — they must be simulated together with cooling, electromagnetic interference, and structural strength.

← Previous Contents Next →
Chapter 114 Duke-Energy

In 1904, James Buchanan Duke — the American tobacco tycoon — founded a small hydroelectric company in North Carolina, harnessing the Catawba River to generate power. His plants supplied electricity to textile mills and tobacco drying furnaces. A hundred and twenty years later, Duke Energy is one of the largest regulated electric utilities in America — providing natural gas and regulated electricity to more than 8 million customers across Florida, Indiana, Ohio, and Kentucky. Its market cap is near one hundred billion dollars.

Not a technological breakthrough. Not a renewable-energy disruption. Not an oil and gas empire.
It was "regulated, stable power supply." Duke is slowly transitioning through the coal-to-gas shift and nuclear plant retirements — accelerating solar in some states, remaining natural-gas-heavy in others. It has survived a century not because it was the fastest — but because "the demand for electricity will never disappear."

Three things:
First: The tobacco tycoon crossed over — using the money from selling cigarettes to build hydropower, transferring capital from lungs to the electrical grid.
Second: A regulated utility is essentially a government-franchised stable return — a principle that has remained unchanged for a century.
Third: The compounding challenge of coal-to-gas transition plus nuclear retirements — Duke must renegotiate the "reliable, cheap, clean" trilemma with regulators year after year.

← Previous Contents Next →
Chapter 115 HCA-Healthcare

1968, Nashville. Dr. Thomas Frist, his father, and a physician partner founded HCA — the Hospital Corporation of America. They built a modern community hospital "managed by doctors themselves" — using standardized supply chains, shared services, and scale purchasing to make surgery and hospitalization cheaper and more consistent in quality than what independent community hospitals could deliver. Today HCA operates more than 180 hospitals across 20 states, with a market cap exceeding one hundred billion dollars.

Not pharmaceuticals. Not insurance. Not diagnostic equipment. Not telemedicine.
It was "using standardized operations thinking to manage bed turnover, surgical scheduling, and care quality across every hospital." The Frist family built a hospital-management industrial cluster in Nashville — a city that later became known as America's "for-profit hospital capital."

Three things:
First: Hospitals can be standardized too — surgical suite utilization rates, average length of stay, and nursing ratios can all be optimized.
Second: Scale purchasing — buying MRI and CT machines on a national scale, driving per-unit costs down to levels that independent hospitals cannot touch.
Third: Trauma services and acute care cannot be done remotely — the hospital is "a physical presence that cannot be outsourced."

← Previous Contents Next →
Chapter 116 Cummins

1919, Columbus, Indiana. Clessie Cummins — a self-taught mechanic and engine fanatic — began installing diesel engines into trucks, funded by his boss. At the time, diesel engines were stationary industrial equipment — used for pumping water and generating electricity. In 1929, during the Great Depression, Cummins drove a truck fitted with a homemade diesel engine across America — no refueling, no gasoline — running on diesel from Indiana all the way to New York. Today Cummins is one of the world's leading manufacturers of diesel and alternative power systems.

Not an auto company. Not gasoline. Not railroads. Not batteries.
It was "diesel + natural gas + hydrogen + electric — any kind of power capable of driving a heavy truck, a mining haul truck, a bulldozer, or a generator." Cummins turned the diesel engine into "a modular heart adaptable to every type of heavy industrial mobile equipment." And now it is putting hydrogen fuel cells and fully electric powertrains into those same chassis.

Three things:
First: It doesn't build trucks — it builds "the heart and lungs of every heavy truck."
Second: In the depths of the Great Depression — crossing America without refueling, proving that a diesel engine could be independent of the gas station network. Extreme marketing = proving it with a physical demonstration.
Third: From diesel to hydrogen — in the decarbonization era, an engine company must become "a power-systems company agnostic to energy medium."

← Previous Contents Next →
Chapter 117 General-Dynamics

1899, Quincy, Massachusetts. An Irish-American inventor named John Holland built the U.S. Navy's first practical submarine — the Holland VI — beside the Fore River Shipyard. His company, the Electric Boat Company, was the predecessor of General Dynamics. Today, General Dynamics is the manufacturer of the M1 Abrams main battle tank, the Virginia-class nuclear submarine, and Gulfstream business jets. Its market cap is near one hundred billion dollars.

Not aircraft carriers (that's Huntington Ingalls). Not fighter jets (that's Lockheed and Boeing). Not missiles (that's Raytheon and Northrop).
It was "tanks and nuclear submarines and Gulfstream private jets." General Dynamics's three pillars — Combat Systems, Marine Systems, and Aerospace — bind three entirely different platforms, spanning armored warfare, submarine warfare, and business aviation from the World War II era to the present, under a single corporate holding.

Three things:
First: From the first practical submarine — to every American nuclear submarine. The continuous carrier of a century of undersea warfare history.
Second: The M1 Abrams's gas turbine engine — not a diesel, but adapted from an aircraft engine. A tank that can run near-silent and on multiple fuel types.
Third: Gulfstream — turning military-grade airframe engineering into the business jet of choice for billionaires and heads of state around the world.

← Previous Contents Next →
Chapter 118 ServiceNow

In 2004, Fred Luddy crawled out of the wreckage of Peregrine Systems. His former company had collapsed due to massive financial fraud. In a small office in San Diego, he built a new company — ServiceNow. The core idea: every workflow in enterprise IT — from employee onboarding and account provisioning to cybersecurity incident response — should not be accomplished by "sending emails" and "manually editing Excel." He turned all of it into automated workflows. Today ServiceNow's market cap exceeds 100 billion dollars.

Not CRM. Not ERP. Not traditional ITSM.
It was "turning every approval, notification, and work dispatch that flows manually through an enterprise — into cloud-based automated tickets." In the early days, Luddy poured nearly all his energy into one thing — a cloud-based IT service desk. The one thing every company needed but none wanted to build themselves. Then he expanded from IT into HR, customer service, and security operations — turning every act of "emailing another department" inside an enterprise into a program.

Three things:
First: It didn't invent new processes — it automated the processes humans had already been running with email and Excel for two decades.
Second: The lesson picked up from the rubble of a bankrupt predecessor — "Never, ever lie on the financial statements. But make absolutely sure the operational processes are transparent."
Third: Platformization — when IT, HR, and security all run on the same automated workflow engine, the switching cost becomes extraordinarily high.

← Previous Contents Next →
Chapter 119 Marriott-International

1927, Washington, D.C. J. Willard Marriott and his wife Alice opened a nine-seat roadside stand — an A&W root beer shop. He had noticed that people in the summer heat couldn't get a cold drink while sitting in their cars — and he invented the concept of "roadside cold refreshments." Later, he expanded it into the Hot Shoppe restaurant chain, and then in 1957 opened the first motel. Today, Marriott International is the largest hotel group in the world, with more than 30 brands ranging from The Ritz-Carlton to Courtyard to Moxy.

Not real estate. Not a tech platform. Not an online travel agency. Not shared accommodation.
It was "giving a good bed and a hot shower to people away from home." Marriott does not own the buildings of most of its hotels — it uses a "management contract" model: real estate investors construct the building and hold the title, while Marriott manages and operates the hotel, collecting management fees and brand licensing. This asset-light model allows it to expand its brands infinitely without requiring infinite capital investment.

Three things:
First: From an A&W root beer stand to 30 hotel brands — but what never changed was "giving a clean bed to people away from home."
Second: The asset-light management contract — carrying none of the capital debt of hotel buildings, yet collecting management profit.
Third: Old Mr. Marriott, well into his nineties, would still personally visit each hotel to measure the fold of the bedsheet corners and the temperature of the coffee.

← Previous Contents Next →
Chapter 120 KKR-&-Co.

In 1976, three former Bear Stearns executives — Jerome Kohlberg, Henry Kravis, and George Roberts — founded KKR in New York. They invented a tool called the "leveraged buyout" — using borrowed money to buy undervalued public companies in their entirety, take them private, fix the operations, and then relist or sell them years later. In 1988, KKR swallowed RJR Nabisco for 25 billion dollars — a deal that has remained an icon in the history of leveraged buyouts ever since. Today, KKR's market cap exceeds 110 billion dollars.

Not a hedge fund. Not a mutual fund. Not venture capital. Not an investment bank.
It was "buying the entire company, fixing it with management and capital restructuring, and then extracting returns from the improved profits." KKR's global investments now span private equity, infrastructure, real estate, and credit. It is no longer a "vulture" — it is "a global holding platform that uses long-term capital to fix and operate a vast array of infrastructure assets and consumer brands."

Three things:
First: An LBO isn't "robbery" — it's "if you're certain a company is worth more, you buy it with other people's money, and then you personally prove it really is worth more."
Second: The RJR Nabisco case was written up as Barbarians at the Gate — but the barbarians later became the operating template for every mega-fund.
Third: When your fund can hold a power generation company for 10 years and restructure its capital — short-term stock price fluctuations cease to be the enemy.

← Previous Contents Next →
Chapter 121 Freeport-McMoRan

1912, Texas. A small mining outfit called Freeport Sulfur Company was founded. Later, it discovered the Grasberg deposit in the remote mountains of New Guinea — one of the largest single copper-and-gold ore bodies in the world. Then it merged with McMoRan to become Freeport-McMoRan. Today it controls some of the richest copper and molybdenum reserves on the planet.

Not tech.
Not finance.
Not oil and gas.
It was copper — a narrative similar to Southern Copper and Newmont, yet distinct: Freeport's core story is Grasberg — a deposit tucked beneath a thirty-thousand-foot peak on the island of Papua — mined across generations in extreme weather, brutal terrain, and an endless back-and-forth with indigenous communities and the Indonesian government.

Three things.
First: Grasberg is "the turquoise-green oxidized veins you can see with the naked eye near the snow line, on top of a copper deposit laced with gold" — and then humanity built the most complex aerial tramway ever devised to haul that ore down the mountain.
Second: Copper is the electrification metal. Whoever wins, copper wins.
Third: Mining political risk only goes so high — when your mine sits on the single largest body of copper reserves on Earth, every government negotiates with you, but none can bypass you.

← Previous Contents Next →
Chapter 122 BNY-Mellon

1784, Alexander Hamilton founded the Bank of New York — the oldest bank in the United States and the first stock ever traded on the New York Stock Exchange. More than two centuries later, it merged with Mellon Financial to become BNY Mellon. Today it is one of the largest custodian banks in the world — safeguarding nearly fifty trillion dollars in custody assets. Every mutual fund, every foreign government bond, every underlying security in every pension account — sits on BNY Mellon's fiduciary ledgers. It is not the flashiest bank — it is the financial industry's great central vault.

Not investment banking. Not lending. Not credit cards. Not asset management.
It was custody — securities safekeeping, asset servicing, clearing and settlement, and corporate trust. When BlackRock buys a share of Apple — BNY Mellon records that share under the correct account name, ensuring that dividends, voting rights, and corporate actions are transmitted correctly.

Three things.
First: Hamilton himself was the architect of America's financial system. This bank has been running the public ledger since birth.
Second: Custody is not about holding money — it is the ultimate registry of who owns every security in the world.
Third: Others may not see you — but if your ledgers are wrong, the entire financial system freezes instantly.

← Previous Contents Next →
Chapter 123 Williams-Companies

1908, brothers Dave and Miller Williams founded the Williams Companies in Arkansas — starting out building oil and gas pipelines. Today it is one of the largest interstate natural gas pipeline operators in the United States — stretching from the Gulf of Mexico to New England, cutting through forests, crossing rivers, buried underground. The natural gas that heats your home in winter, that generates your electricity in summer — roughly one-third of it travels through Williams' pipelines.

Not extraction. Not refining. Not power generation. Not sales.
It was the pipeline — moving natural gas from the wellhead all the way to the city gate station. Gas pipelines, like the electrical grid, are equally invisible and equally critical to modern society. Williams' Transco pipeline system is the single most important natural gas transport artery on the U.S. East Coast.

Three things.
First: Once a pipeline is buried and the rights-of-way are locked in — no competitor can run parallel, because no one can dig a second parallel trench across the same private land.
Second: The shale revolution sent U.S. natural gas production into overdrive — every molecule of that new supply must move out by pipeline.
Third: Natural gas is the transition fuel — until wind, solar, and storage are fully mature, gas-fired power is the baseload replacing coal.

← Previous Contents Next →
Chapter 124 Comcast

1963, Ralph Roberts bought a small cable TV system with just 1,200 subscribers in Tupelo, Mississippi. That was the seed of Comcast. Later, his son Brian Roberts transformed that little cable company into America's largest broadband and cable television media group — acquiring NBCUniversal, Universal Studios, and Sky. Today Comcast has a market cap of roughly $150 billion.

Not telecom. Not satellite. Not a tech startup. Not original content.
It was bundling the pipes and the programming together — broadband is the conduit, NBC and Universal are the content. Over two generations, the Roberts family turned a tiny cable system into a media empire that shapes global entertainment and internet. They never sold — "cable plus content" is the family creed.

Three things.
First: It started with a small-town cable system serving 1,200 subscribers, then snowballed into a giant.
Second: Vertical integration — owning both the broadband pipes and the content factory — locking the data stream and the entertainment stream inside the same corporate account.
Third: Broadband has become an essential utility, like water and electricity — and then the broadband bill becomes a fixed monthly tax.

← Previous Contents Next →
Chapter 125 FedEx

1965, Yale University. Fred Smith proposed an "overnight express" air network model in a term paper that his professor graded a C — all packages would fly to a single central sorting hub, then be redistributed in the dead of night onto flights heading to their destinations. The professor wrote in the margin: "This idea is not feasible. Because you would be competing with the U.S. Postal Service."

1971, Smith took that paper — the one that had been graded a failure — along with the few million dollars he had inherited and $90 million in financing, and founded Federal Express in Memphis. On the first night, April 17, 1973, fourteen Falcon jets took off from cities across the country, carrying 186 packages bound for the Memphis sorting hub. Today the purple-and-orange FedEx logo is ubiquitous on airport tarmacs around the globe.

Not the post office. Not freight. Not passenger aviation. Not warehousing.
It was turning Memphis into the midnight revolving door for every package in America — Fred Smith engineered logistics into a hub-and-spoke network. Before express delivery, a package had to go from A directly to B; long detours were prohibitively expensive. Smith's insight: fly every package to a single midpoint first — spend nearly the entire night auto-sorting them there — then dispatch them out before dawn. Mathematically, the total path length is shorter than point-to-point.

Three things.
First: A Yale professor's "C" and "you're competing with the U.S. Postal Service" — turned into one of the most successful logistics companies in the world.
Second: The core is mathematical shortest total path — every package only needs to be sorted once to go from any A to any B.
Third: FedEx jets fly across every continent in the early morning hours — in the internet age, the speed at which physical objects move is still the physical bottleneck of e-commerce.

← Previous Contents Next →
Chapter 126 Intercontinental-Exchange

2000, Jeffrey Sprecher — a serial entrepreneur who started out in power trading — founded the Intercontinental Exchange (ICE) in Atlanta. It began as an online energy trading platform. Then he executed a series of audacious acquisitions — including the New York Stock Exchange (NYSE) itself. Today ICE has a market cap exceeding one hundred billion dollars, and nearly every interest rate, equity, energy derivative, and mortgage data point you encounter has passed through an ICE data center and trading engine.

Not a brokerage. Not a bank. Not an investment group. Not an algorithmic fund.
It was the holy trinity of exchange + clearing + data. Sprecher's logic: control the physical venue where prices form, then control the benchmark data, and then customers have to pay to see prices they themselves generated. ICE started with energy, then swallowed NYBOT (coffee, sugar, cocoa), then swallowed the NYSE — turning "exchange data" into recurring subscription revenue far more valuable than trading fees.

Three things.
First: Buying the New York Stock Exchange — letting a company born in 2000 from an energy platform become the owner of the very symbol of American capitalism.
Second: Trading fees fluctuate with hot and cold markets — data subscriptions are stable, SaaS-style recurring revenue every single month.
Third: Control the place where prices form — and you can sell both the trades and the data about the trades.

← Previous Contents Next →
Chapter 127 McKesson

1833, New York City. John McKesson and Charles Olcott formed a wholesale drug importing partnership in Lower Manhattan — bringing in bottles of herbs, tinctures, and opium produced in Europe and selling them to American doctors and pharmacists. Nearly two hundred years later, McKesson is one of the largest pharmaceutical distributors and healthcare supply chain companies in America — moving every box of prescription medication from Pfizer's and Merck's factories to the pharmacy counter at every CVS and Walmart.

Not drug manufacturing. Not the pharmacy. Not diagnostics. Not the hospital.
It was the logistics layer of medicine — McKesson's trucks and automated distribution centers crisscross the American landscape carrying prescription drugs the way hemoglobin carries oxygen through the bloodstream. In the most intense period of COVID-19 vaccine distribution — McKesson was one of the centralized distributors designated by the U.S. government for vaccines and ancillary supplies.

Three things.
First: They don't make the drugs — they make sure that every pharmacy, when it opens its doors tomorrow morning, has on its shelves the medicine that was missing today.
Second: The pharmaceutical supply chain cannot break — a break means patients go without medicine. And a drug shortage is a political scandal at best, medically fatal at worst.
Third: Two hundred years, from importing European herbs to distributing mRNA vaccines — yet the role has never changed: the intermediary in the supply chain.

← Previous Contents Next →
Chapter 128 Waste-Management

1968, Dean Buntrock started a small garbage company in Chicago — collect trash, then bury it. Then in 1971, he merged it with several local waste haulers to form Waste Management. It did something nobody found glamorous — buying up landfills across America, painting its garbage trucks a uniform green. Today WM is the largest waste management and environmental services company in the United States, with a market cap approaching one hundred billion dollars. What it controls is not technology — it is the place your garbage has to go.

Not tech. Not recycling technology. Not eco-friendly materials. Not circular-economy products.
It was the final destination of trash. WM's landfills and transfer stations form a web of physical monopoly across America — because getting a new landfill built today is extraordinarily difficult politically, environmentally, and in terms of land scarcity. When everyone else has to send their garbage to WM's landfills every day — WM doesn't charge by the month. It charges by the ton.

Three things.
First: A landfill is the ultimate physical moat — because no one wants a brand-new landfill next to their home.
Second: Turning solid-waste methane into natural gas — recovering value from the decomposing organic matter in the landfill.
Third: The essence of waste management isn't recycling — it is that humanity generates garbage every single day, and WM is the one who has to be called that very night.

← Previous Contents Next →
Chapter 129 ADP

1949, New Jersey. Henry Taub was an accountant. He noticed that every company spent enormous amounts of manual labor calculating wages and tax withholdings before each payday — a purely manual process that was inefficient and maddeningly error-prone. He founded a company called "Automatic Data Processing," renting IBM punch-card machines to do payroll calculations for other companies. ADP went on to become the world's largest payroll and human resources outsourcing company.

Not HR software. Not a recruiting agency. Not consulting. Not hiring.
It was the machine behind the moment every single person gets paid each month — the one calculating tax withholdings, Social Security, and health insurance deductions. ADP handles the payroll for roughly one in every six private-sector employees in America. Its data is an exquisitely sensitive macroeconomic leading indicator — the Bureau of Labor Statistics' nonfarm payrolls report relies in part on ADP data.

Three things.
First: It didn't start as an "HR platform" — it started by automating the excruciatingly painful weekly manual labor of payroll calculation.
Second: Payroll cannot be wrong — miscalculate a single tax point, and apologies are meaningless. That means high switching costs and a high trust threshold.
Third: Aggregating anonymized payroll data and selling it to macroeconomists — ADP became the private satellite tracking the American job market.

← Previous Contents Next →
Chapter 130 PNC-Financial-Services

1845, Pittsburgh. Pittsburgh Trust Company opened in a small office, providing trust administration and local bank lending for the fortunes of the era's steel and railroad magnates. Through a long series of mergers — most notably with National City Corporation, absorbing substantial assets in the process — it became today's PNC Financial Services Group. PNC is the quintessential super-regional bank — with a dense network of branches and corporate clients across the Midwest and Mid-Atlantic.

Not a Wall Street investment bank. Not a global consumer bank. Not a payments network giant.
It was the biggest local bank in the great cities of the Midwest. PNC is woven into the fabric of business and consumer life in Pittsburgh and Philadelphia — loans, payroll, mortgages, wealth management. It is a conservative, regulated, deposit-funded bank — often far safer in a crisis than a portfolio of investment-banking bets.

Three things.
First: The financial bloodstream of an industrial city — as steel and manufacturing declined and healthcare and education rose, PNC regenerated alongside the city itself.
Second: It never chased global expansion — it went deep instead of wide in the states it already occupied.
Third: The super-regional bank model — large enough to have an efficient technology platform, small enough to know the name of the best mortgage client on every Main Street.

← Previous Contents Next →
Chapter 131 Monster-Beverage

1935, California. A family juice workshop run by a man named Hubert Hansen began selling fresh juice to the movie studios of Southern California. Decades later, his grandsons Rodney Sacks and Hilton Schlosberg took an unremarkable energy drink brand — Monster — and turned it into a feral black beast with green claws. Today Monster Beverage has a market cap approaching one hundred billion dollars. Its green-claw logo fights an endless refrigerated street brawl with Red Bull inside every gas station cooler on the planet.

Not Coca-Cola. Not Pepsi. Not coffee. Not sports drinks.
It was the energy drink — caffeine, taurine, B vitamins — married to a visual identity of extreme sports plus heavy metal plus monster. Monster didn't compete by tasting better than Red Bull — it tattooed itself beneath the skin of adrenaline-soaked youth through heavy sponsorship of the X Games, esports, and mud-track motocross.

Three things.
First: It didn't change the formula philosophy — it changed the cultural definition of who was drinking it. Red Bull is F1; Monster is dirt bikes and esports.
Second: In 2015, Coca-Cola bought roughly 16.7% of Monster — granting it access to the Coca-Cola distribution system, which overnight gave it a global retail pipeline.
Third: It's not selling a beverage — it's selling the legitimization of the energy to code through midnight, skateboard at dawn, and race off-road by morning.

← Previous Contents Next →
Chapter 132 Fortinet

2000, Sunnyvale. Ken Xie — a Tsinghua-educated cybersecurity engineer — founded Fortinet. His previous company, NetScreen, had been acquired by Juniper. He didn't retire. He set out to build a cheaper next-generation firewall — accelerated by custom ASIC chips designed in-house. Today Fortinet is worth tens of billions of dollars, with millions of FortiGate firewalls and endpoint devices online around the world. It is one of Palo Alto Networks' fiercest competitors in this segment.

Not antivirus. Not VPN. Not penetration testing. Not SIEM.
It was cybersecurity accelerated by hardware chips — faster, cheaper, but just as secure. Ken Xie himself came out of chip engineering — he offloaded the firewall's most CPU-intensive task, deep packet inspection, onto a dedicated custom chip, delivering application-layer security inspection without throttling throughput.

Three things.
First: Not pure software — it's a security ASIC chip running inside a hardware firewall.
Second: Fortinet's strategy is packing next-generation firewall capabilities into an extremely low price point — so mid-sized enterprises can afford them.
Third: Ken Xie's background at the intersection of security engineering and semiconductors — allowing Fortinet to compete on two tracks simultaneously: silicon plus code.

← Previous Contents Next →
Chapter 133 U.S.-Bancorp

1863, the same month Lincoln signed the National Bank Act, First National Bank of the United States was founded in Minneapolis — the direct predecessor of today's U.S. Bancorp. A hundred and sixty years later, U.S. Bancorp has thousands of branches across the Midwest and West, with trillions in trust and payments operations. It is one of the five largest banks in the United States by assets.

Not Wall Street. Not international. Not investment-bank risk-taking. Not financial innovation.
It was the everyday deposits, loans, mortgages, and trusts of the Midwest. U.S. Bancorp is renowned for conservative underwriting standards and a remarkably small pool of bad assets. During the subprime crisis, it was one of the only large regional banks that stayed consistently profitable. Why? Because in 2005 it had voluntarily exited the subprime market — not because it foresaw the crisis, but because it felt that "loan terms we ourselves can't understand — we shouldn't be selling to customers."

Three things.
First: If you can't understand the loan, don't sell it — that's a culture, not a compliance clause.
Second: Payments processing and trust — the flank the media barely notices — contributes steady non-interest income.
Third: The "super" in super-regional bank doesn't mean international expansion — it means achieving extraordinarily high penetration in just a handful of states.

← Previous Contents Next →
Chapter 134 Elevance-Health

1944, Indianapolis. Mutual Medical Insurance, the Blue Cross Blue Shield plan for Indiana, was formed — the predecessor to what would become Anthem, and later Elevance Health. Today it is one of the largest health insurance companies in America — providing health benefits, insurance, and care management to more than forty million members. The blue of Blue Cross Blue Shield has existed for eighty years as a federated network of nonprofit and for-profit hybrids stretching across every state.

Not hospitals. Not pharma. Not diagnostics. Not medical devices.
It was the insurance layer that settles the bill between you and the hospital, the pharmacy, and the doctor when you get sick. Elevance (Anthem) serves as the intermediary payer and care manager for employer-sponsored plans and Medicaid/Medicare programs across dozens of states. Its core innovation has been shifting from after-the-fact reimbursement to population health and preventive care.

Three things.
First: Health insurance isn't a bet that you won't get sick — it's managing the probability of illness across an entire population and how early intervention happens.
Second: The rebrand from Anthem to Elevance — symbolically stepping out of the "insurance" label, declaring that the business is about elevating health.
Third: The network effect of the Blue Cross Blue Shield Association — every state has its own Blue plan, but mutual brand recognition means any Blue plan member can walk into any Blue network hospital anywhere in the country.

← Previous Contents Next →
Chapter 135 UPS

1907, Seattle. A nineteen-year-old named Jim Casey, with a borrowed hundred dollars and a secondhand bicycle, started a local messenger service called the American Messenger Company. He rode that bicycle himself in the rain, carrying letters and packages from one office building to another. Today UPS is one of the largest package delivery and logistics companies on Earth — that single bicycle has become hundreds of thousands of brown delivery trucks and hundreds of cargo aircraft. The color brown was Jim Casey's personal choice — "because it looks clean, and you can't see the dirt when it's always running through mud."

Not the post office. Not high-speed rail. Not air cargo. Not ocean shipping.
It was getting the thing you bought from the warehouse to your doorstep — in that brown truck. UPS and FedEx have competed for decades in overnight air and ground delivery — each evolved a mirror-image global sorting-and-distribution system. UPS's deep industrial-engineering culture — drivers open the door with their left hand, the key is always in the right pocket, every route has been mathematically optimized — is the Taylorist philosophy of logistics.

Three things.
First: It started with one bicycle — and step by step, laid down the largest ground parcel delivery network in the world.
Second: The color brown was chosen because "mud splatter won't show" — an early decision of extreme pragmatism that became the deepest imprint of the brand.
Third: Logistics is an industrial-engineering science — shave one second of waiting at a red light on every left turn, and across millions of packages you save millions of gallons of gasoline.

← Previous Contents Next →
Chapter 136 American-Tower

1995, Boston. American Tower was spun out of American Broadcasting Company (one of CBS's predecessor entities). The logic at the time was simple: broadcasters and mobile carriers were scrambling to mount antennas on rooftops and water towers — so why not turn the antenna pole itself into an independent, rentable piece of infrastructure? Today American Tower owns more than 200,000 communications towers and rooftop sites globally — renting space around the world to AT&T, Verizon, T-Mobile, and every carrier that wants to deploy 5G small cells.

Not a telecom carrier. Not a tower manufacturer. Not fiber. Not spectrum.
It was leasing space in the sky. A tower's location is irreplicable — nobody wants a new steel tower built in their backyard. So the owner of a tower that already exists can rent multiple antenna layers in the same physical space to multiple carriers simultaneously.

Three things.
First: Passive infrastructure — once the tower is built, every additional tenant carries a gross margin approaching 100%.
Second: Each generation of wireless, from 3G to 5G — adds another layer of antennas to the same tower, and the rent stacks again.
Third: The physical and political irreplicability of a tower site — worth far more than the steel in the tower itself.

← Previous Contents Next →
Chapter 137 CSX-Corporation

1827, Maryland. The Baltimore and Ohio Railroad — one of the oldest predecessor railways for freight and coal in America — laid its first stone in a ceremony witnessed by Charles Carroll, a signer of the Declaration of Independence. Today CSX operates tens of thousands of miles of freight rail across the eastern United States — hauling Appalachian coal, Midwestern grain, Gulf Coast chemicals, and Southeastern containers all the way to East Coast ports.

Not high-speed rail. Not passenger service. Not trucking. Not pipelines.
It was heavy freight, long distance, steel rail. Rail freight burns far less carbon and energy per ton-mile than trucking — and when your port is buried under mountains of containers, a single standard freight rail car can replace seven or eight semi-trailers. CSX strings together countless factory zones and port terminals across the eastern United States on a line of steel — continuously switching cars through the complex process of American deindustrialization and the reshoring of near-shore manufacturing.

Three things.
First: The rails were laid before the Civil War — the rail network is a geographic skeleton that cannot be replicated.
Second: Shifting freight from trucks to rail is a self-interested choice that automatically serves sustainable logistics — because rail is simply more fuel-efficient.
Third: The locomotive engineer and the dispatcher — not an app — are living human beings at the control stand, coordinating by radio.

← Previous Contents Next →
Chapter 138 Enterprise-Products

1968, Houston. Dan Duncan started Enterprise Products with a single propane truck and a small propane transfer station. Then he laid propane pipelines into, out of, and across the Gulf Coast. Today Enterprise Products Partners is one of the largest midstream energy pipeline and export terminal companies in North America — moving crude oil, natural gas liquids, ethane, propylene, and petrochemical feedstocks from the Permian Basin and Eagle Ford all the way to export terminals on the Gulf Coast, then loaded onto ships bound for the world.

Not upstream extraction. Not refining. Not gas stations. Not trading.
It was the pipeline — the same logic as Williams but positioned in different basins and on different hydrocarbon molecule streams. Enterprise is the most critical carrier of NGL (natural gas liquids) and propylene pipelines — fractionating the wet gas at the shale wellhead into ethane for ethylene crackers, moving propylene to plastics plants. In the first decade of the clean energy transition — global plastics demand is still rising — and ethane and propylene are the precursors to plastic.

Three things.
First: Propane and ethane are not crude oil — but the pipeline logic for them is perfectly parallel.
Second: Midstream assets are toll booths — no matter how upstream prices swing, every barrel that passes through pays a toll.
Third: The Gulf Coast is the global bottleneck for petrochemical exports — and EPD owns the valves on most of the pipelines leading into that bottleneck.

← Previous Contents Next →
Chapter 139 Schlumberger

1926, Alsace, France. Brothers Conrad and Marcel Schlumberger — both geophysicists — invented the first downhole electrical resistivity well-logging cable for oil exploration. They lowered electrodes down a borehole, sent an electric current into the rock formations, and measured the returning waveforms — thereby deducing where oil layers lay dozens of meters underground without ever drilling out a core sample. Their "well logging" would become the foundational technique of all oil exploration. Today Schlumberger is the largest oilfield services company in the world (SLB). The SLB blue logo is visible on drilling rigs and fracturing sites across every continent.

Not an oil company. Not refining. Not trading. Not petrochemicals.
It was every technical service needed to help oil companies find oil — and pull it out from thousands of meters beneath the earth. From well logging, drill bits, drilling fluids, to completions, reservoir modeling, and production optimization — SLB provides end-to-end technical support for exploration and production to the world's major oil companies.

Three things.
First: That logging cable from 1926 — it turned oil exploration from relying on nose and luck into seeing through rock strata with physical instruments.
Second: Oil companies bear the reserve risk; SLB bears the technical services — no matter how oil prices swing, wells still need to be logged and cemented.
Third: The rebranding from Schlumberger to SLB — symbolic of a identity shift from "oil services" to "energy technology services."

← Previous Contents Next →
Chapter 140 SLB-Schlumberger

1912, Alsace, France. Two Schlumberger brothers — Conrad and Marcel — set up a strange device in the barn behind their family home: a wooden box fitted with four copper electrodes, powered by batteries, buried in the earth. They weren't looking for groundwater — they were measuring changes in soil resistivity. Conrad was a physicist; Marcel was a mining engineer. They invented the world's first electrical well-logging apparatus — inferring oil in rock formations by sending electric current through the ground. That method went on to become Schlumberger — today the largest oilfield services company on Earth.

Not a spike in oil prices.
Not a petroleum giant.
Not a drilling permit.
It was two brothers sticking four copper rods into the soil. The principle they discovered was stunningly simple: rock layers at different depths have different electrical conductivities — oil-bearing sandstone has high resistivity, water-bearing shale has low resistivity. If you can measure resistivity segment by segment down the wall of a borehole, you can "see" from the surface where oil sits thousands of meters underground — without having to dig it out first.

Conrad Schlumberger was born in 1878 in Alsace to a textile-manufacturing family — Alsace belonged to Germany at the time. He studied physics at the École Normale Supérieure in Paris — his classmates included Paul Langevin and Pierre Curie. His early research focused on the thermoelectric properties of minerals — but later, as a professor at the Paris School of Mines, he began using mines for electrical exploration experiments. Marcel managed coal mines across southeastern Europe and the Middle East — he showed Conrad just how complex the underground really was.

1927 — the two brothers, driving an old Renault truck loaded with cable spools and a current recorder — performed the world's first resistivity well-logging operation at the Pechelbronn oil field in France. They lowered an electrode probe 500 meters down a borehole on a cable, receiving the resistivity signal at the surface. The result was unmistakable — the resistivity curve precisely pinpointed the oil-bearing sandstone layers. Then they took that curve chart to the oil companies to sell it. The oil companies looked at the chart and asked: "Are you magicians or frauds?"

Not magic. It was Maxwell's equations — the physical law governing how an electric field travels through media of different conductivity. Schlumberger's "magic" was mathematically inverting Maxwell's equations — inferring subsurface structure from surface measurements — a process called inversion. Inversion algorithms remain one of the core mathematical frameworks of geophysics to this day.

By the 1930s, Schlumberger had expanded from France to the United States, Venezuela, the Soviet Union, and the Middle East — because no matter which continent you stand on, the physics underground is the same. Today Schlumberger's market spans well logging, drilling, completions, stimulation, production optimization, carbon sequestration, and geothermal — with operations in over 120 countries.

Three things.
First: Measure with fundamental physics — don't guess.
Before electrical well logging, the oil industry found oil by having geologists look at surface outcrops — which was basically guessing. The Schlumberger brothers didn't offer a better guess — they used Maxwell's equations to replace guessing with physical measurement.
In your industry — which critical decision is still made by human guesswork, when a sensor plus a physics equation could take its place?
Second: Run it again — France to America to the world.
Schlumberger started in the French market, but Alsace had always sat at the cultural crossroads of France and Germany. Under CEO Pierre Schlumberger (Conrad's son), the company moved its headquarters from Paris to Houston in 1940 — not out of love for America, but because the Nazis were coming.
Your globalization — is it truly adapting to every market — or is the physical principle itself universally applicable?
Third: Oilfield services isn't about selling equipment — it's about selling the certainty of knowing what lies underground.
Oil companies pay Schlumberger for measurement fees, interpretation fees, stimulation design fees — not because the equipment is expensive, but because without Schlumberger's data, drilling a dry hole costs upwards of fifty million dollars.
Is your product priced against the cost of the customer making the wrong decision — or against your own costs?

Conrad Schlumberger died in 1936, of a stroke, while traveling to Sweden to deliver a geophysics lecture. Marcel continued running the company until 1940. The two brothers corresponded in French their entire lives — the core subject of their letters was always: can the mathematical inversion of the resistivity curve be made more stable? They never discussed money in those letters.

← Previous Contents Next →
Chapter 141 Airbnb

October 2007, San Francisco. Two Rhode Island School of Design graduates — Brian Chesky and Joe Gebbia — couldn't pay their rent. They noticed that a design conference in town had booked every hotel room solid. They pulled out three air mattresses, set them up in their living room, and put up a listing: "Air mattress + lodging + breakfast — $80." Then they emailed every conference attendee who couldn't find a hotel. Three people showed up. That weekend they made $1,000 — just enough to cover the rent gap while Chesky was thousands of dollars in credit card debt. Today Airbnb is worth over $100 billion.

Not the sharing economy. Not a travel recovery. Not anything that existed before.
Just three people sleeping on air mattresses for one night. Chesky and Gebbia's landlord — a computer engineer named Nathan Blecharczyk — lived in the same apartment. Blecharczyk later became the third co-founder and CTO. The three of them built the first version of Airbedandbreakfast.com — with exactly three listings: the three air mattresses in their apartment living room.

In 2008 AirBed&Breakfast went to the DEMO conference — and was publicly ridiculed. Paul Graham, one of Silicon Valley's most famous investors (and founder of Y Combinator), gave them this verdict: "Are you out of your minds? Who would stay at a stranger's house?" But Graham still accepted them into YC — because he saw something else: during the 2008 election, the three founders designed and hand-packaged 1,000 boxes of limited-edition "Obama O's" and "Cap'n McCain's" breakfast cereal — packaging and mailing every box themselves — just to pay their credit card bills. Graham said: "If they can sell cereal — they can sell houses."

Then Airbnb spent four years doing something everyone said was "impossible" — building trust between strangers. They first required hosts to upload real photos of themselves. Then they mandated linked credit cards and government ID. Then they launched a two-way review system (hosts rate guests, guests rate hosts), and up to $1 million in host property insurance. These trust infrastructure pieces weren't "nice-to-haves" — they were "without them — the platform would have been shut down after a San Francisco apartment was ransacked by a guest in 2011."

Airbnb was hit with a sledgehammer during the 2020 pandemic — global travel volume dropped to near zero. The company's valuation fell from $31 billion to $18 billion, and Chesky laid off 1,900 people — roughly 25% of the workforce — apologizing in an open letter at the same time. Then, within that same year — the remote-working population started using Airbnb in place of short-term rentals, and Q3 2020 profits actually turned positive. At the end of 2020 Airbnb went public at a $100 billion valuation — more than doubling on its first day. In 2022, Chesky announced a permanent remote work policy. "We don't require employees to come to the office — because our product is in every house in the world."

Three things.
First: the best startups grow out of the founder's own survival needs.
Chesky and Gebbia didn't "identify a market opportunity" — they couldn't pay their rent. They were solving their own survival problem.
Could the pain you're going through right now — become a product you can sell?
Second: don't wait for others to believe in you — build the trust machine yourself.
The idea of strangers staying in strangers' houses — ridiculed by skeptics for four years — became normal through three pieces of infrastructure: identity verification, two-way reviews, and $1 million insurance.
The core premise of your business that people doubt most — can you eliminate that doubt with engineering, rather than persuasion?
Third: the worst moments expose what the real core asset is.
In April 2020 Airbnb's daily bookings were cut by 80%, but by summer 2020 — during city lockdowns, people started using Airbnb to stay a month at a time within a two-hour drive of their cities. Chesky said: "The pandemic told us — Airbnb's core was never travel. It was 'living anywhere.'"

Brian Chesky still hasn't bought his own house. He rents in San Francisco — booking his own places on Airbnb. He says: "The moment I don't use my own product — I lose the right to talk to hosts."

← Previous Contents Next →
Chapter 142 Boston-Scientific

1979, Boston. Two medical salesmen — John Abele and Pete Nicholas — sat in a Dunkin' Donuts next to Massachusetts General Hospital. Abele had just quit another interventional medical device company — they didn't want his coronary balloon angioplasty catheter design. On a napkin he drew a diagram of a balloon inside a blood vessel, being inflated by liquid pressure. Nicholas looked at it and said, "Wouldn't that thing burst the vessel wall if you put it in?" Abele said: "It won't — because the balloon uses polyethylene terephthalate, PET — it locks at a preset diameter and doesn't go further." They finished their coffee — and founded Boston Scientific. Today the company is worth over $120 billion.

Not a medical breakthrough. Not surgery. Not basic research.
Just taking an inflatable plastic balloon — machining it to micron-level precision — threading it along a guidewire into a blocked coronary artery, inflating it with saline, and compressing the plaque. This procedure is now called percutaneous transluminal coronary angioplasty, or PTCA — performed millions of times a year in hospitals around the world.

John Abele was born in 1937 in Ohio, earned his undergraduate degree at Boston College and his MBA at Harvard. He'd been a sales manager at a medical device subsidiary under Johnson & Johnson. Shadowing doctors in hospitals, he watched them crack open entire chest cavities for bypass surgery on blocked coronary arteries — the mortality rate was not trivial, and recovery was measured in months. Abele started working in a small lab in the Boston area — together with an engineer named Robert L. Martin — hand-molding polyester balloons using precision thermal forming. The earliest molds were cut from aluminum on a lathe in a garage.

Boston Scientific's first breakout wasn't Abele's balloon itself. It was his integrated "guide catheter + balloon catheter + pressure pump" delivery system — which allowed a single interventional cardiologist to perform the entire balloon dilation procedure alone. Before this, every interventional procedure required three people working in coordination — an interventional radiologist, a cardiologist, and a scrub nurse — because the devices weren't compatible with each other.

In the 1990s, Boston Scientific expanded its portfolio through continuous acquisitions — coronary stents, peripheral vascular stents, cardiac electrophysiology electrodes, endoscopic hemostatic clips and sphincterotomes — gradually transforming from "that balloon company" into "full-spectrum minimally invasive interventional devices."

In 2006, Boston Scientific executed what Harvard Business School would later write up as a case study in "the most dangerous acquisition" — a $27 billion hostile takeover of cardiac rhythm management giant Guidant. With extreme leverage. Afterwards, because of Guidant product recalls and Boston Scientific's own plunging stock price, the company carried what may have been the heaviest debt load in medical device history for a period. But ultimately — it brought the leverage down using the positive cash flow from its pain management and cardiovascular businesses.

Three things.
First: just make one balloon — but its precision has to be locked to the inner diameter of the vessel.
Every early R&D effort at Boston Scientific revolved around a single physical parameter — the balloon's inflated diameter had to be accurate to within 0.25 millimeters. Any wider — the vessel wall tears. Any narrower — the plaque doesn't expand, and the artery re-occludes within two days.
Is your product's core physical-parameter safety tolerance narrow enough to earn life-or-death trust?
Second: what competitors dismiss — might be your entire fortune.
Abele pitched the balloon idea inside J&J — and was rejected. Then he took the same rejected draft — and built a business within his own company that reached tens of billions in annual revenue.
That proposal of yours that got shot down — where is it now?
Third: a hostile acquisition nearly destroyed the company — and then it saved itself.
Guidant was a hostile acquisition by Boston Scientific in 2006 — saddling it with a mountain of short-term debt and ICD pacemaker recall penalties, and the stock price halved and halved again. Then it spent nearly a decade paying down the debt. You can never avoid making one wrong big bet — but you can hold on and wait for a medical market's positive cash flow to slowly plug the bullet holes.

John Abele still goes to work every day — in his 80s. He doesn't have a fancy office — he sits at a cubicle desk at headquarters in Marlborough, Massachusetts, surrounded by engineers. His motto is: "Better is better."

← Previous Contents Next →
Chapter 143 Moodys

1900, Lower Manhattan, New York. A young man who had spent years working as a newsboy and securities statistician on Wall Street — John Moody — sat in a narrow office without heating, drawing grids by hand. He was making something that every Wall Street bank considered "utterly worthless": a small handbook compiling the assets, liabilities, earnings, and management of every railroad company in America. On the cover he wrote a label in his own handwriting: "Moody's Manual of Railroad Investments." Every word of the first edition was written out by hand — then he went and sold copies himself at newsstands next to train stations and outside brokerage offices. The Moody's that followed — one of the three global credit rating agencies — is worth roughly $80 billion today.

Not a regulatory license. Not a financial instrument. Not a bond bull market.
Just one man copying the financial reports of every railroad company by hand.

John Moody was born in 1868 to an ordinary family in New York. At 17 he began working on Wall Street as a securities statistician — which meant manually recording the annual report numbers of listed companies in lined notebooks. In those days there was no SEC, no disclosure standards, no audits — every railroad company published its accounts using its own accounting conventions. Some treated maintenance costs as capital expenditures; others hid debt outside the consolidated statements of subsidiaries.

Moody discovered a crack — investors (especially British and European investors wanting to buy American railroad bonds) had no way of knowing which railroad company's liabilities were real and which were fabricated. So he decided to do one thing — manually standardize the balance sheets and income statements of every major railroad in America into a uniform format, and assign a simple credit rating from A to E. His first manual was published in 1900 — and sold out within the month. In 1909 he extended railroad credit ratings to industrial bonds — and in 1914 he founded Moody's Investors Service.

Moody's credit rating mechanism wasn't about "predicting the future" — it was about "standardizing public information so investors could compare for themselves." But because that standardization was so useful — it eventually got written into U.S. securities law in the 1930s and later into global bank capital accords. The Basel Accords mandated that banks use rating agency credit grades to calculate risk-weighted capital — turning Moody's, S&P, and Fitch into the "unofficial infrastructure" of the global financial system.

The 2008 financial crisis — Moody's credit ratings were pushed to the center of universal attack. It had rated hundreds of billions of dollars' worth of securities packed with subprime loans as AAA — the highest level of safety. And those securities suffered widespread defaults over the following two years. At a congressional hearing, one senator told Moody's CEO: "You are like a market inspector who labels rotten fish as fresh."

Moody's defense was that they were rating "a statistical probability of credit risk measurement," not offering "a 100% guarantee." But under legal and public pressure, the ratings industry underwent a complete restructuring of its methodologies and board governance.

Three things.
First: standardization — is itself a gold mine.
Moody didn't invent any data. He took existing data — railroad annual reports — and standardized it. That act of standardization made everyone believe "comparison is possible" — then "comparable" became essential — then "essential" became a rating license written into law.
In your industry — is there an information flow that "everyone handles but no one formats the same way" — waiting to be standardized?
Second: a rating is not a prophecy — it's a confidence interval.
Moody's ratings are not "absolute safety" — they are a ranking of historical default probability. But what the 2008 crisis exposed was that investors were treating "statistical probability" as if it were "a government guarantee."
Are your customers amplifying your stated meaning in the way they use your judgment?
Third: when your rating becomes infrastructure — you are no longer a company.
Moody's credit ratings were written into global banking regulations. That means every rating decision Moody's makes is simultaneously an act of regulatory compliance. This kind of power is extraordinarily rare in business history — but the accompanying "cannot be wrong" responsibility is also more than ordinary people can bear.
Has your product been written into a certain rule — turning you into an "irreplaceable bolt"?

John Moody died in 1955 — and in the months before his death, he still went to the office every day to read periodic railroad bond reports. In his entire life, he never once said "I got this one right" — only "which company did I miss."

← Previous Contents Next →
Chapter 144 Mondelez

1898, Chicago. A wholesale clerk at a small grocery — Fred E. Stephen — bought the exclusive U.S. distribution rights to Tobler Swiss milk chocolate from a traveling salesman who had just returned from Switzerland. Then he added: Milka, Lu, and the parent company of Oreo, Nabisco (acquired as part of Kraft in 1999). In 2012, Kraft Foods spun off its global snack business into a new company called "Mondelēz International" — a name forged by welding together words for "delicious" and "world" from different languages on earth. Today Mondelēz is worth over $90 billion, and Oreo cookies alone generate more than $4 billion in annual sales.

Not food technology. Not snack trends. Not health-conscious consumption.
It's that motion you made as a kid — twisting open an Oreo, licking off the cream, then dunking the dark cookie in milk. Mondelēz turned this little biscuit, born in 1912 at a bakery in Chelsea, New York, into the most recognizable cookie on the planet. A century of "twist-lick-dunk" — globally consistent.

But the real story behind it is this: Mondelēz was split off from Kraft Foods in 2012 — separating the high-margin but slow-growth "North American grocery" business (Kraft macaroni, cheese, salad dressing) from the lower-margin but faster-growth global "snacks" business (Oreo, Cadbury, Toblerone, Ritz crackers). The logic for the latter: emerging-market middle-class expansion → double-digit annual growth in packaged snack consumption → need a global distribution machine focused solely on "sweet and salty."

From the moment of the split, Mondelēz CEO Dirk Van de Put did one thing — shift all growth into emerging markets. In India, Mondelēz priced single Oreo cookies at 5 rupees (less than half a U.S. cent equivalent in purchasing power at the time), penetrating hundreds of thousands of village mom-and-pop shops — delivering double-digit compound annual sales growth for a decade straight. In China, Oreo was remade in a chameleon coat of flavors ranging from green tea to mustard. In Brazil, Tang powdered drink mix was rebuilt into the best-selling powdered beverage — because every local household had cold water and a refrigerator, just no chilled fresh juice.

Three things.
First: a spin-off — is about releasing two businesses with different "time horizons" from being each other's prison.
Kraft grocery: stable, slow, high-margin, North America. Snacks: fast, volatile, global, high-growth. Same board, same balance sheet — the grocery side sucked up the snack side's investment. Once separated — each side's decision clock sped up or slowed down on its own, no longer dragging the other down.
Do you have a "fast" business — currently being managed by the management of a "slow" business?
Second: global brand — local metamorphosis.
Oreo's flavor differences across China, India, and Brazil are so extreme you'd barely recognize them as the same cookie. But the "twist-lick-dunk" motion works in every market — because it's a physical action, not language.
Could your global brand's core anchor be "a motion" rather than "a slogan"?
Third: selling for pocket change — can capture the widest base of the global pyramid more effectively than a "premium positioning."
A single Oreo sold for 5 rupees in India — while Mondelēz's competitors were still selling imported cookie rolls for 20 rupees. Mondelēz wasn't selling "cheaper cookies" — it was selling "the first industrialized sweet experience in an Indian child's life."
Can your pricing go low enough to make your product into someone's "first taste"?

The name "Mondelēz" itself is a creation — "monde" (French: world) + "deliz" (Italian: delicious). In 2012, some brand team at Kraft used this assembled word to re-baptize a century-old snack company — like hanging a brand-new neon sign on an old cathedral.

← Previous Contents Next →
Chapter 145 Ciena

1992, Maryland. Three engineers who had worked on fiber optic and military communications at General Instrument — David Huber, Kevin Haines, and Patrick Nettles — quit their jobs and secured their first $2 million from a venture capital fund. Their idea sounded insane at the time: use lasers of 16 different color wavelengths — transmitted simultaneously over a single optical fiber, with each wavelength independently carrying its own set of signals. This approach later came to be called Wavelength Division Multiplexing, or WDM — the physical foundation for scaling up the internet backbone. Today Ciena is worth roughly $20 billion.

Not fiber. Not the internet. Not cloud computing.
Just taking a single strand of glass and turning it from "one lane" into "16 parallel lanes" — and later 160, then 320 lanes. Ciena's founders weren't the first to conceive of WDM — but they were the ones who used silicon fabrication processes to take wavelength division multiplexing from "a laser rack in a lab" to "a single line card that a telecom company could plug into a standard equipment rack."

Before joining Ciena, Patrick Nettles had been a network engineer at British Telecom and GTE — he knew the telecom operators' sharpest pain point: every few hundred kilometers on a long-haul backbone, the optical signal had to be converted into an electrical signal for regeneration — each regenerator was a cabinet-sized piece of equipment, enormously power-hungry, astronomically expensive, and it hard-capped the speed of bandwidth upgrades. Ciena's solution: no regenerators — use Erbium-Doped Fiber Amplifiers, or EDFAs, to amplify the optical signal directly in the optical layer, skipping the "optical-electrical-optical conversion" entirely.

In 1996, Ciena sold its first 16-wavelength WDM system to Sprint — instantly multiplying the bandwidth of a single fiber by 16 times. Then WorldCom, AT&T, and MCI all became customers — because internet traffic was doubling every 12 months, yet the cost of laying new fiber and the environmental permitting cycle simply couldn't keep pace. WDM became the magic that liberated "fiber already buried in the ground" — no new cable laid, using colored light to paint dozens of times more capacity.

The dot-com bubble burst in 2000 — and Ciena, along with the entire all-optical networking industry, was decimated. Ciena's stock plunged from over $1,000 (pre-split adjusted) to a few dozen dollars. But it survived — by extending WDM technology into metro fiber networks and submarine cable projects, and by continuously acquiring self-healing optical network software companies — connecting the optical layer end to end.

Three things.
First: physical bottlenecks are the greatest business opportunity.
Fiber isn't expensive — burying fiber is expensive. What Ciena saw was that "already-buried fiber was being utilized at under 5%" — and it used WDM to turn one strand into dozens of virtual strands. Physical scarcity → an engineering breakthrough in frequency multiplexing → billions in annual revenue.
In your industry — are there "physical facilities that already exist but are utilized at an extremely low rate" — waiting for a frequency multiplexing solution?
Second: it wasn't the people who invented WDM who won — it was the people who crammed WDM into a 19-inch rack unit.
Bell Labs had described WDM in papers back in the 1980s. What Ciena did: use their own silicon photonic integrated circuits, self-developed wavelength-locked lasers, and standardized network management software — and turn it into a telecom product.
Is your technology — in a paper or in a product?
Third: a company's prospects can shift in an instant because of the slope of the network traffic curve.
In the late 1990s, internet traffic doubled every 12 months — and the doubling cycle of Ciena's WDM capacity passively resonated with that slope. They didn't "predict success" — they sat on the correct side of an exponential curve.
Which curve are you sitting on?

Patrick Nettles retired from Ciena in 2008. He nearly vanished along with the company after the dot-com crash of 2001 — but he and Huber pivoted Ciena in 2002 from pure backbone optical transmission into metro and edge — and that adjustment saved the company. He said: "Light doesn't tire — light only gets obstructed. A network engineer's job is to move the obstacles out of the photon's path."

← Previous Contents Next →
Chapter 146 Marsh-McLennan

1871, Chicago. After the Great Chicago Fire burned down half the city, Henry W. Marsh made a decision — he would no longer be just a railroad insurance broker. He would open a company that "put all of a company's industrial risks — from fire to boiler explosion to marine shipping — on a single sheet of paper and calculated the premium." That was the beginning of Marsh. In 1905, Alexander McLennan founded a separate insurance actuarial and employee benefits consulting firm in Boston. The two merged in 1906. Today Marsh & McLennan is worth roughly $80 billion, one of the largest insurance brokerage and professional services firms in the world.

Not an insurance company. Not reinsurance. Not actuarial formulas. Not a legal mandate.
Just two men who saw the same thing: corporations — especially large industrial corporations — had no idea how much risk they carried, nor any ability to negotiate with insurers. Marsh and McLennan bundled together four things — risk identification, coverage actuarial calculation, insurance bidding, and claims management — and sold the package to every company that had a boiler, a fleet of ships, or a factory floor.

Henry Marsh was born in 1848 on a farm in Illinois. As a teenager he hauled freight at Chicago's train stations, then entered the railroad insurance trade — dealing daily with railroad company treasurers, selling fire and accident insurance. In 1871 the Great Chicago Fire incinerated most of the city — many insurance companies went bankrupt and fled outright, and policies turned to scrap paper. From that fire Marsh learned a truth that every insurance broker would repeat for the next 150 years: "When you buy insurance, you're not just buying a policy — you're buying the insurance company's balance sheet."

So he began doing a double check: examining the financial health of insurance companies for his clients — while simultaneously examining his clients' own risk inventories. At the time, boiler explosions at American factories were happening roughly once every two days — Marsh hired retired engineers from boiler manufacturers and shipyards as "risk surveyors," sent to client sites to find the cracks.

Over the same period, McLennan made a different decision: he turned employee benefits and pension actuarial science — two concepts that were extraordinarily rare in 1905 — into a consulting service. After World War II, Marsh & McLennan acquired its way into Mercer Human Resources Consulting (one of the largest HR management consultancies in the world), Oliver Wyman (one of the top ten global strategy consulting firms), and Kroll — binding together four pillars: insurance + actuarial + strategy + human resources consulting.

Three things.
First: insurance isn't about buying a product — it's about buying the knowledge of "what can go wrong."
The diagnosis issued by Marsh's risk surveyors was often: "this boiler weld joint will fail from fatigue cracking within three years." An insurance company couldn't price that risk — so Marsh stood in the middle and delivered the full suite: risk recommendations + coverage design + competitive bidding across multiple insurers.
Is the value of your intermediation — "matching" or "diagnosing"?
Second: turn engineers into insurance consultants — don't turn salespeople into consultants.
Marsh hired retired boilermakers and shipwrights to inspect factory safety. Mercer Human Resources Consulting hired statisticians and labor lawyers to do pension actuarial work. In this industry, the real moat of the intermediary is the asymmetry of specialized expertise — not client relationships.
Is your company hiring "salespeople" or "engineers with diagnostic capability"?
Third: the merger wasn't because "we've got enough" — it was because the two sides' skill trees were fully complementary and entirely non-overlapping.
Marsh handled property and casualty insurance and liability — McLennan handled employee benefits and actuarial work. After the merger, their clients didn't say "I'm buying different services from two different places" — they said "finally, one person I can entrust with all my risks."

Henry Marsh retired in 1915. All his life he insisted that "every policy must be checked by my own hands before it goes to the client." His company would later grow larger than the world's biggest insurance carriers — but he never once entered the direct underwriting business. He said: "A broker and an underwriter — are like a lawyer and a judge. You can only choose one role."

← Previous Contents Next →
Chapter 147 Bloom-Energy

2001, NASA Ames Research Center. An Indian-American scientist named Dr. K.R. Sridhar had just designed a solid oxide fuel cell device for NASA — one that could produce oxygen from carbon dioxide and hydrogen in the Martian atmosphere. That device operated successfully in a simulated Mars environment. When Sridhar came back to Earth, he reversed the design direction of that device — not to produce oxygen, but to generate electricity. Using natural gas, biogas, and hydrogen to generate power directly through a solid oxide fuel cell, with no combustion. Today Bloom Energy is worth roughly $25 billion.

Not clean-energy policy. Not carbon taxes. Not grid modernization. Not geothermal.
A byproduct of NASA's Mars program. Sridhar was the former director of space technology at NASA's Ames Research Center — his team had been working on "in-situ resource utilization for Mars" in the late 1990s — how to use the CO₂ in the Martian atmosphere to produce oxygen for astronaut life support and rocket fuel. The core device was a high-temperature solid oxide electrolyzer capable of running in reverse.

Sridhar realized that if this ceramic-electrolyte fuel cell were run in the opposite direction — with natural gas and air as input — it could produce electricity directly, at higher efficiency than combustion-based generation, and without the nitrogen oxides and sulfur oxides that burning produces. He founded Bloom Energy in 2001 — and went silent for eight years. No public product. No publicity.

In 2010, the Bloom Energy Server — a solid oxide fuel cell unit roughly the size of a parking space — was shown to the outside world for the first time. Each Bloom Box generates about 200–300 kilowatts of electricity. The earliest major customers were Google, eBay, FedEx, and Walmart — they didn't buy Bloom Boxes for green PR. They bought them because California's electricity is expensive and subject to frequent blackouts, and the per-kilowatt-hour cost of a Bloom Box was lower than the grid price.

Bloom Energy went public in 2018. At the company's IPO, Sridhar wore an old NASA badge — he said it was to remind himself: the physical foundation of this company came from a space experiment about "trying to survive on an alien world."

Three things.
First: a survival device for Mars — brought back to Earth to become a grid replacement.
Something NASA designed for the Martian atmosphere — Sridhar asked one reverse question: "Why not run it the other way in our living rooms?" He didn't invent new physical principles — he reversed the direction of a reaction.
In your industry — is there a physical or chemical process that's only ever been used in one direction — that could be reversed into something new?
Second: endure eight years — without revealing a single product detail.
Bloom Energy maintained total secrecy from 2001 to 2009 — no media, no published performance data, only internal testing at a handful of major Silicon Valley companies. Sridhar's logic: solid oxide fuel cells were a technology that had existed for decades. If you reveal too early, the noise you generate activates every incumbent — and they'll crush you on price and distribution channels before your product matures.
Can your new thing — right now — withstand an all-out counterattack from incumbents?
Third: sell electricity — not "green."
The number-one reason Bloom Energy's customers bought it wasn't environmental — it was "cheap + stable." California's grid electricity prices swing wildly, and summers bring frequent outages. The Bloom Box's cost per kilowatt-hour was lower than the grid's peak price — so Silicon Valley data centers bought them the way they'd buy backup generators — except this thing runs 24 hours a day.
Can your product sell on "price-performance" — rather than on "moral virtue"?

Sridhar still signs off on technical drawings every day. He says: "Generating electricity on Earth and producing oxygen on Mars are governed by the same thermodynamic equation — only the boundary conditions changed."

← Previous Contents Next →
Chapter 148 Northrop-Grumman

1930, Inglewood, California. A Norwegian-American engineer who had come out of a Los Angeles aircraft factory built a shed-like workshop at the edge of the desert. His name was Jack Northrop. For his favorite plane — the Alpha — he chose an all-aluminum monoplane with an all-metal stressed-skin design, at a time when every other aircraft was still a biplane of canvas and wood. He made the fuselage into a single seamless aluminum shell — no struts, no bracing wires, rivets driven from the inside out. The Alpha's cruising speed was faster than the military fighters of its day — not because of a big engine, but because of low aerodynamic drag. Today Northrop Grumman is the manufacturer of the world's most advanced stealth bomber, the B-2 Spirit, and is worth roughly $80 billion.

Not military budgets. Not geopolitics. Not industrial policy.
Just Jack Northrop's lifelong obsession with a single aerodynamic configuration — the flying wing — from the first sketch he drew in 1929 all the way to the B-2 stealth bomber's first flight in 1989.

Jack Northrop was born in 1895 in Newark — the descendant of Norwegian immigrants. He went straight from high school into Lockheed, where he was the company's most brilliant aircraft structural designer before Kelly Johnson came along. He designed the fuselage of the Lockheed Vega — the aircraft Amelia Earhart would later fly across the Atlantic.

But Northrop's obsession was not with conventional aircraft — it was with the flying wing. A flying wing: no fuselage, no horizontal tail, the entire aircraft is a single wing. The ultimate aerodynamic form — minimum drag, maximum lift, and extremely unstable. In 1940 he built the N-1M test aircraft — a pure flying wing, with no vertical control surfaces at all, steering by differential drag at the wingtips. The test pilot said handling the thing felt like "balancing a clothes iron on the palm of your hand and walking with it."

He later built the XB-35 flying-wing bomber — canceled by the Air Force because the jet age had arrived. He converted it into the jet-powered YB-49 — canceled again over stability problems. By the time he saw the B-2, he was in the final stages of liver cancer — in a hospital bed. The Air Force specially arranged for an officer to carry a B-2 model and its blueprints into his hospital room. After seeing it, he said: "Now I know why God has kept me alive for 25 years." He died eight months later.

In 1994 Northrop Corporation merged with Grumman — the latter being the contractor behind the F-14 Tomcat fighter, the EA-6B Prowler electronic warfare aircraft, the lunar module, and the Apollo program. After the merger, Northrop Grumman became one of the five giants of American military aerospace — the B-2, the Global Hawk drone, the E-2D Hawkeye early warning aircraft, and the future B-21 Raider stealth bomber — all of them flying wings.

Three things.
First: hold fast to an aerodynamic vision — for 60 years.
From the first flying-wing sketch Northrop drew in 1929 to seeing the B-2 in 1989 — 60 years. His company collapsed, restructured, merged, and was rejected by the Air Force three times — but he never once abandoned the flying-wing configuration.
Do you have a principled vision — worth clinging to for 60 years?
Second: the most beautiful aerodynamic form — is also the hardest to control.
A flying wing achieves perfect drag and lift — but with no tail, it's profoundly unstable. The B-2 was ultimately able to fly not because of structural improvements — but because fly-by-wire flight computers could adjust control surface deflection hundreds of times per second. Northrop's physical intuition was correct in the 1940s — but the control systems didn't catch up to his imagination until the 1980s.
Is your intuition — waiting for a "digital control system" to mature?
Third: seeing the B-2 model from a hospital bed — that image in itself is the definition of an engineering legacy.
Northrop didn't live to see the B-2 take flight. But he saw the blueprints — and then he died. His designs were carried forward by others, his company was acquired by another company, but his flying-wing principle ultimately became the highest form of America's air-based strategic deterrence.
Can you accept — that your work will reach its apex after you are gone?

Jack Northrop died in 1981. His office was preserved after his death — on the drafting board by the window, still pinned there, was an unfinished charcoal sketch of a flying wing's tip. In the lower right corner, in his handwriting, was the annotation: "Drag = 0.0013..."

← Previous Contents Next →
Chapter 149 Simon-Property-Group

1960, Indianapolis. A young man named Melvin Simon — freshly discharged from the military — used the pay he'd saved up during the Korean War, together with his brother Herb, to build his first open-air shopping plaza on a cornfield just outside Indianapolis. He wasn't a real estate developer — he was a mortgage loan broker. His reasoning for building the plaza was brutally simple: in the default records he handled, commercial retail defaulted at far lower rates than residential — because a residential tenant could just stop paying rent and move out, but a grocery store had to stay in the same location, since losing regular customers during a move would cost even more. Today Simon Property Group is the largest owner of shopping malls and outlet centers in the world, worth roughly $60 billion.

Not commercial real estate cycles. Not a retail revolution. Not e-commerce.
Just the hidden pattern that "commercial retail defaults at a lower rate" — discovered in a stack of loan default files.

Melvin Simon was born in 1926 in the Bronx, New York — to a Jewish immigrant family. He earned a business degree at City College of New York under the GI Bill, then worked as a mortgage loan broker in Indianapolis — sitting in an office every day reviewing loan applications and studying borrower credit reports. He split the residential loan data and the commercial loan data into two piles and compared them year by year — and found that commercial retail loan loss rates during economic downturns were only one-quarter of residential. The reason was simple: a store can't move — moving means losing your regular customers, and the renovation costs for a new space would be a total write-off.

In 1960 the Simon Property Group (in its predecessor form) was founded. By the 1970s Simon had laid down over a dozen neighborhood shopping centers across Indiana and surrounding states — and then he made a decisive pivot: no more neighborhood shopping centers. He started building regional malls. He treated national department store chains like JCPenney and Sears as "anchor tenants" — offering them rock-bottom rent, even zero, in exchange for 20-year leases and the foot traffic they'd draw. Then he filled the rest of the mall around these anchors — leasing to all the smaller retailers at far higher rental rates and pocketing the spread.

In the 1990s Simon went on an acquisition rampage — swallowing the DeBartolo Group in 1996, then engulfing the mall chain giant Corporate Property Investors in 1998, becoming America's largest mall owner overnight. Then he created the Premium Outlets brand — using the "discount luxury" concept to pull consumers into destination shopping villages an hour's drive into the suburbs.

The e-commerce onslaught — Amazon drove JCPenney and Sears into waves of store closures. After 2016, Simon pivoted its anchor tenants from traditional department stores to gyms, movie theaters, dining experiences, and offline showrooms for online-native brands — transforming the mall from "a place to buy things" into "a place to eat, work out, and watch movies."

Three things.
First: in default files, find an asset class safer than residential.
Melvin Simon's insight wasn't in financial statements — it was in the corner of a default-rate statistics table. He discovered a fact overlooked by every real estate investor: the physical cost of a tenant relocating, plus customer inertia, makes commercial retail leases more stable during recessions than residential leases.
In your industry — is there a form of low risk created by a "hidden cost" — that everyone has missed?
Second: anchor rent can be zero — as long as the anchor brings enough people.
Simon exchanged "free or zero rent" for Macy's and JCPenney signing 20-year leases — then used that foot traffic to stuff 200 small tenants in at high rents.
Do you have an anchor tenant you'd accept zero profit on just to hold your place in the ecosystem — because its presence exponentially amplifies your margins everywhere else?
Third: a mall doesn't sell goods — it sells "a reason to leave the house."
During the decade e-commerce was at its most ferocious — Simon restructured the mall from department-store-dominated to a mix of "dining + fitness + cinema + pop-up shops." People can buy clothes online — but they can't eat at a restaurant online, can't rock-climb online, can't meet friends in a cinema lobby online.
Is what the internet is replacing in your product the "function" or the "experience"? If it's the experience — can your digital competitors outperform you on that experience?

Melvin Simon died in 2009. Before handing the CEO role to his eldest son David Simon in 1993, he never drove past a mall and judged whether it "looked good" — he evaluated one thing and one thing only: "Pedestrian density at the storefronts. Stand there with a stopwatch and a counter for ten minutes — and you'll know whether this place has value."

← Previous Contents Next →
Chapter 150 Monolithic-Power-Systems

1997, Los Gatos, California. A former engineer who had spent over a decade designing analog power chips at Texas Instruments and ADI — Michael Hsing — began designing a new DC-DC converter chip at his dining room table. At the time, the analog power chip sector was an arena absolutely dominated by a handful of major players: Texas Instruments, Linear Technology, Maxim, and Power Integrations. Hsing's idea was to use a BCD (Bipolar-CMOS-DMOS) process to place high-voltage power devices and low-voltage control logic on the same piece of silicon — thereby creating a "fully integrated power conversion module" — rather than just a standalone voltage regulator. Today MPS is worth roughly $25 billion.

Not semiconductor cycles. Not power electronics. Not Chinese demand.
Just putting high-voltage devices and a low-voltage controller onto the same piece of silicon.

Michael Hsing was born in Shanghai, earned his bachelor's degree at Zhejiang University, then went to the U.S. for his Ph.D. He first worked as a power circuit designer at ADI — and discovered that the entire power management chip industry was trapped inside a physical box: the high-voltage power-stage MOSFET had to sit on a separate chip — because standard CMOS processes couldn't withstand 30V to 100V. Meanwhile the low-voltage control chip sat on a separate piece of CMOS. The two chips were connected by bond wires — introducing parasitic inductance, increasing size, and degrading efficiency.

Hsing applied the BCD process — a technology that can simultaneously fabricate Bipolar (precision analog), CMOS (low-power logic), and DMOS (high-voltage power) on a single piece of silicon — to power chips. MPS's first product was an ultra-small 3A step-down converter with efficiency exceeding 95% — something no one else could do at the time. It required no external MOSFET switch — everything was integrated into one package.

Physically, this level of integration meant "the power module can be made small enough to tuck right next to a laptop, an LED driver, automotive electronics, and — later — the explosively growing population of data center GPUs." Every NVIDIA data center GPU needs dozens of power rails alongside it — and MPS's compact, high-efficiency power modules happened to be a perfect match for the massive surge in AI data center power demand.

MPS went public in 2004 — and unlike many chip companies, it didn't explode overnight. It climbed to the tens-of-billions level with a compound annual growth rate of 15% to 25% per year. Because the analog power chip market isn't a "blockbuster" market — it's a "steady penetration" market. Today nearly every cloud server, every electric vehicle, every PC motherboard has MPS products inside.

Three things.
First: turn two chips into one — physical integration is the competitive moat.
MPS's competitors can build a voltage regulator with the same specs — but they have to use two pieces of silicon plus bond wires. MPS's single-silicon solution is inherently smaller, cheaper, and more reliable — because the bond wire count is cut in half directly. This isn't "cleverer design" — at the device physics level, an entire packaging layer has been eliminated.
Can you defeat your competitors — by "removing one physical component" — instead of by "offering a lower price"?
Second: the slow company — analog chips don't fear slowness.
It took MPS 27 years from a bedroom to a $25 billion market cap — with an average annual growth rate under 20%. But the analog chip market isn't "explode and die" — it's "once you're designed in, the customer's switching cost is higher than replacing their CEO."
Can your industry tolerate "slow but irreplaceable"?
Third: become the "hidden infrastructure" of a category.
Nobody says "my computer uses an MPS power chip" — but every machine depends on one. MPS is a textbook example of "zero brand, one hundred percent penetration."
Are you willing to let your brand disappear — in exchange for every user being physically dependent on you?

Michael Hsing remains CEO of MPS to this day. He rarely gives public talks or interviews. Among his engineers in Silicon Valley, the motto that gets repeated is: "MPS doesn't stand for a name — it stands for Monolithic. Put everything onto a single chip — then say nothing."

← Previous Contents Next →
Chapter 151 Cigna

1792, Philadelphia. America's first marine insurance company — the Insurance Company of North America, or INA — was established by a charter from the Pennsylvania legislature. Its first president was Dr. Benjamin Rush, one of the signers of the Declaration of Independence. But the real core was what it would go on to underwrite — from cargo losses on slave trade ships (the 1790s), to gold shipments during the California Gold Rush (1849), to the aviation insurance on TWA's first Boeing 707 crossing the Atlantic (the 1960s). In 1982, INA merged with Connecticut General Life Insurance Company — and took the name Cigna. Today Cigna is worth roughly $90 billion, one of America's largest employee health insurance and pharmacy benefit management companies.

Not healthcare legislation. Not longevity trends. Not government contracts.
Just multiple rebirths spanning 230 years — from marine insurance to fire insurance to employee health benefits and global healthcare services.

Ralph Saul, Cigna's CEO in the 1970s during the INA era, made a breakthrough decision — he refused to turn the company into a giant all-lines composite insurer. Instead, he concentrated INA on the high-technical-complexity segment of commercial property and casualty insurance. Around the same time, he introduced the concept of "risk management" to corporate clients for the first time — not selling policies, but helping large companies like DuPont and GM design captive insurance programs and reinsurance structures. This was the transformation from "insurer" to "risk advisor."

The 1982 merger with Connecticut General — on paper, the rationale was cross-selling (property-and-casualty clients need life insurance, life insurance clients need P&C). The real chemistry didn't emerge until after the 1990s: Cigna discovered that its two pillars — employee benefits (corporate health insurance, dental, disability, life insurance) and "health services" (pharmacy benefit management, medical networks, telehealth) — were more stable than property-and-casualty, and delivered double-digit compound growth. So it gradually divested its P&C and catastrophe reinsurance operations and shifted its center of gravity entirely onto "the human body."

Then came international expansion — Cigna built its own wholly-owned health insurance and TPA (third-party administrator) networks in select Asian and Middle Eastern countries, directly transplanting the American model of employer-paid health benefits. This strategy proved especially successful in Singapore, the UAE, and Saudi Arabia — because local governments wanted to shift the medical costs of expatriate white-collar populations from government-run universal healthcare onto employer-purchased private insurance.

Three things.
First: an insurance company doesn't have to cover everything — but what you cover must be the thing with the highest "probability of occurrence + severity."
Cigna dropped auto insurance, homeowners insurance, and catastrophe reinsurance — and concentrated on "the money a person needs between falling ill and recovering" and "the health services a person needs after retirement." The average premium on this single policy is far higher than auto insurance — and the insured person's depth of suffering is far deeper — which means a far lower inclination to comparison-shop.
What you're covering — is it your customer's "dandruff" or "heart failure"? Depth of suffering = pricing power.
Second: don't compete with government healthcare — cover what government healthcare doesn't reach.
In many countries, Cigna complements rather than competes with government universal healthcare — focusing on premium private hospitals, expedited diagnostics, and international medical referral services. In every country on earth with a middle-class population in the millions or above, there is a small segment of people willing to pay three times the price for the right to see a specialist within three hours.
In your market — is there a gap where "public system supply is inadequate + private willingness to pay is extremely high"?
Third: a health insurance company's greatest asset is not policies — it's "medical behavior data."
Cigna manages the health consultations, pharmacy records, and surgical data of tens of millions of people. With that data — it can predict in advance the probability that a diabetic patient will be hospitalized within the next two years, and dispatch someone to intervene ahead of time. This data capability is a moat that no one else can access.

Cigna announced in 2024 that current CEO David Cordani will retire after 15 years at the helm. He said the biggest mistake of his tenure was "not exiting property-and-casualty entirely sooner." He keeps in his possession to this day a photocopy of INA's original 1792 parchment charter — the original sits in an insurance museum.

← Previous Contents Next →
Chapter 152 Sherwin-Williams

1866, Cleveland. Henry Sherwin — the son of a bookseller from rural Ohio — bought the inventory and formula notes of a failed paint shop with $2,000. He opened every can of paint in stock, measured the ingredient ratios in each — and discovered that three batches of the same white paint from the same supplier had titanium dioxide content that varied by 40%. He closed all the cans and said to his sole partner: "We need to do one thing — buy a mixer." His partner asked why. Sherwin said: "Paint has to be the same ratio, every time." Today Sherwin-Williams is worth roughly $80 billion, the largest paint company in the world.

Not a chemical breakthrough. Not color science. Not real estate cycles.
Just "three batches with different titanium content" — an industrial norm that every painter accepted as given. In 1866, paint was mixed by general stores according to secret formulas — every shop turned out something different. What color a consumer ended up with was basically a matter of luck. Sherwin introduced weighing and mechanical mixing into paint production — standardized the formula — and turned color repeatability into a product.

Henry Sherwin was born in 1842 in Ohio; his family sold books. He dropped out of college halfway through and worked in wholesale groceries in Cleveland. In 1866, he happened to walk past a failed little paint shop and saw the mountains of pigment powder, lead white, linseed oil, and turpentine drums piled inside — the owner had been stirring paint with a broom, and a hardened crust often built up at the bottom of each barrel.

The first thing Sherwin did after buying the shop wasn't to sell more — it was to buy a precision balance scale from a pharmacy, reweigh every formula, and write it all down. Then he went to a blacksmith and commissioned a custom "mechanical mixer" — something no one in the paint industry used at the time — essentially a large iron drum with a spiral blade fixed to a rotating shaft. He named the device "The Mixer." Before the Mixer, all paint was stirred by hand; inconsistent mixing times meant inconsistent colors. The Mixer enforced a constant rotation speed and a fixed mixing time.

Then Sherwin pushed forward on three fronts simultaneously: developing quick-drying paint (because hand-mixed paint dried slowly — after painting, you had to wait days before moving the furniture back in); building company-owned retail stores (skipping the wholesalers, so a painter could walk straight into a shop with a Sherwin-Williams sign and buy paint and tinting paste); and a color standardization system (launching color cards and formula catalogs — so the same color code painted in New York and San Francisco would come out exactly the same).

The 1920s — Sherwin-Williams invented the first nitrocellulose lacquer (Duco) for General Motors — cutting automobile painting time from two weeks down to a few hours. The 1950s — introduced the first water-based latex paint. Today Sherwin-Williams has over 5,000 company-owned stores worldwide — and has acquired over a dozen brands, including Valspar.

Three things.
First: sell "consistency" to everyone.
Before Sherwin, everyone buying paint knew the color would be off — because that was an established fact of the industry. He took something that already existed and made it unacceptable — then eliminated the deviation with a mixer.
In your industry — is there a "quality deviation everyone accepts" — that could be eliminated with a "mixer"?
Second: company-owned stores — don't be a wholesaler.
Sherwin built his own stores early on — because a wholesale clerk couldn't explain to a painter how to match colors. But a company store salesperson could — they were painters themselves, trained in color formulation.
Do you distribute — through "people unrelated to you" or through "specialists you trained yourself"?
Third: from painting walls to painting cars — the same physical skill, translated sideways.
Latex paint became nitrocellulose lacquer — change the solvent, change the resin, change the curing agent. But Sherwin learned this: the essence of a coating is adjusting the mixture ratio of "pigment particles + binder + solvent" so it adheres uniformly to any surface. This core capability translates sideways onto walls, cars, aircraft, oil tanks, ship hulls, pipe interiors.

Henry Sherwin died in 1916. In his will he left one line: "I am to be buried at Lake View Cemetery in Cleveland. Please do not paint my headstone — let the rain fall on it."

← Previous Contents Next →
Chapter 153 Lumentum

2015, Milpitas, California. An optical networking equipment company called JDSU — after nearly two decades of struggle — performed a surgical split: it spun off its optical communications laser and photonics business as Lumentum, and turned its network testing and measurement business into Viavi. The day the separation papers were signed, Lumentum CEO Alan Lowe wrote a single line in an internal email: "At last, we only do light." Today Lumentum is worth roughly $12 billion, and its lasers — particularly the VCSEL lasers used in data center fiber optic transceiver modules and Apple's iPhone 3D sensing — are among the most advanced commercial semiconductor laser sources in the world.

Not a demand explosion. Not 5G. Not 3D facial recognition.
Just the recognition that "semiconductor lasers as an independent category — need to be carved out of a big, everything-and-the-kitchen-sink optical testing company."

JDSU had been one of the hottest stocks on Wall Street during the 1999 internet bubble — because it simultaneously did optical components, fiber optic test instruments, optical amplifiers, and optical switching systems. When the bubble burst, the stock fell 99%, and for the next fourteen or fifteen years the company existed in a state of "neither growing nor dying." JDSU's management and board eventually admitted a simple fact: optical communications components and optical test instruments — are two different sets of customers, two different R&D cycles, two different profit-margin models. Put together — one drags down the other.

Immediately after the spin-off, Lumentum invested aggressively in VCSELs (Vertical-Cavity Surface-Emitting Lasers) — miniature semiconductor lasers that can be mass-produced at extremely low cost. In 2017, Apple integrated Face ID for the first time in the iPhone X — with a VCSEL array projecting tens of thousands of invisible infrared dots onto the human face to generate a depth map. Lumentum was Apple's primary VCSEL supplier. That single decision sent its revenue and profits soaring in just 18 months.

At the same time, on the optical communications side, Lumentum supplies hyperscale data centers (AWS, Google, Microsoft Azure) with high-speed tunable lasers and coherent optical transceiver modules, supporting 400G and 800G fiber connections. The demand for optical interconnects inside AI data centers is growing exponentially — and Lumentum is one of the few American optical component manufacturers in this market capable of providing the full chain from "laser to optical module to amplification."

Three things.
First: the spin-off — is to keep a good piece of meat from being smothered in the fat of another piece.
JDSU was a tossed salad — Lumentum was the shrimp buried in it. Pulled out, the shrimp alone is worth $12 billion. Left in the salad, the shrimp is drowned in dressing and you can't taste it.
Does your company — have a piece of "shrimp"?
Second: inside your biggest customer's physical product — become an irreplaceable layer.
Apple's Face ID has a VCSEL array inside every iPhone. The cost of that array is negligible relative to the iPhone's total price — but if it fails, the phone's primary biometric authentication function is dead. Lumentum doesn't "sell lasers to Apple" — it is embedded in Apple's core hardware architecture.
To your customer — are you a "supplier" or "a layer embedded in their architecture"?
Third: bet on a single process — VCSEL — and then wait for its application scenarios to keep expanding after you're gone.
VCSELs went first into 3D facial recognition. Then automotive LiDAR. Then short-range optical communications. Then possibly AR/VR headset micro-projectors. A single process — infinite downstream applications.
Do you have a "process" — that can be translated sideways, indefinitely, into different downstream markets?

Alan Lowe retired from Lumentum in 2022. Among the employee email replies he received on the day of the spin-off, there was one from a production line operator — a person who had worked at JDSU for 18 years. His exact words were: "Eighteen years — and this is the first time I've known what my product is called."

← Previous Contents Next →
Chapter 154 OReilly-Automotive

1957, Springfield, Missouri. Charles F. O'Reilly and his son Charles H. O'Reilly — an auto mechanic back from World War II — rented an abandoned gas station shed off Route 66 on the edge of town and opened the first O'Reilly Auto Parts store. On opening day, their shelves held exactly two products: spark plugs and radiator belts. They couldn't afford more inventory. They sold both items to a truck driver who had been changing his own tires for 16 years but had never once swapped a spark plug himself. Today O'Reilly Automotive is worth roughly $80 billion.

Not car ownership rates. Not DIY culture. Not chain retail.
Just a father and son, starting with a truck driver's one flickering-yellow spark plug — and building a single inventory system that consolidates the parts needs of every "fix it yourself" person and every professional repair shop in America.

Charles H. O'Reilly (known as "Chub" O'Reilly) had served in the Navy in the Pacific — repairing engines on PT boats and landing craft. Back in Missouri, he worked as a service technician at a local car dealership. He observed a strange phenomenon: chain parts stores like Napa and Carquest primarily served professional repair shops — and turned away small orders from regular people. If an ordinary person wanted to swap a spark plug or a wiper blade on a Saturday afternoon — they either went to Sears (limited selection) or a small gas station (only carrying their own brand).

Chub O'Reilly did two small things, small for the time and the place: first, he opened the store's door to everyday residents — whether or not you were a professional repair shop, if you came in to buy a one-dollar spark plug, he'd personally look up your vehicle model for you; second, he positioned his stores in county seats, rural towns, and suburbs — not big cities — because when a country person's car breaks down, there's no Uber to call, and "fix it yourself or get the neighbor to help" is the only option.

O'Reilly went public in 1993 — then began a massive wave of acquiring small regional auto parts chains and folding them all into a single inventory management system. O'Reilly's SKU count far exceeds that of a typical auto parts store — its warehouse system means a customer can walk into a store in a rural town of just three thousand people, and the clerk can have the part delivered from the regional distribution hub by the next day — at the same price.

Three things.
First: serve the small customers that big companies ignore — then watch "small" become "large."
The customer base at O'Reilly's first store was individual DIYers — a group that both Napa and Carquest had ignored. At the time, the auto parts world believed "small, scattered retail customers aren't worth the trouble." O'Reilly spent decades proving that — in aggregate — scattered small customers are in fact the largest segment of auto parts consumption.
In your industry — who is "that small customer group" that big companies are actively giving up on?
Second: on Route 66 — not in a big city.
O'Reilly chose the rural route — opening stores in county seats, farming towns, and suburbs. In these places, rent is low, competition is zero, and the customer's brand-switching radius is extremely high (because this is the only store nearby).
When you open a store — do you choose "downtown" or "the only one within a 50-mile radius"?
Third: the core of the parts business isn't the parts — it's "having exactly the one you need."
Inventory breadth. Inventory predictability. Regional replenishment speed. These three things work on a logic different from retail — it's logistics infrastructure. Back in the 1990s, O'Reilly built its own distribution center network and automated replenishment algorithms to predict the time window when repair shops in small towns start stocking up on antifreeze for winter.

Charles F. O'Reilly died in 1972. His grandson David O'Reilly later became CEO — and between 2008 and 2018 he led the company's market cap to grow more than tenfold. He said his grandfather left behind only one work notebook from the first store, and on one page, it read: "This customer says the spark plug threads don't match his Ford. Check whether Ford changed the thread spec or we mislabeled our inventory number."

← Previous Contents Next →
Chapter 155 Emerson

1890, St. Louis, Missouri. Judge John Wesley Emerson — the grandson of a Scottish-American immigrant — opened a small electric motor factory on the banks of the Mississippi River. He manufactured only one thing: electric motors powered by the alternating current from Niagara Falls. At the time, Edison's direct-current grid was locked in a battle with the AC grid of Westinghouse and Tesla for control of America's electrical standard. Emerson bet on AC — he wrote a letter to Tesla himself, asking for detailed advice on AC motor design. Tesla wrote back. Today Emerson is worth roughly $75 billion, one of the world's largest suppliers of industrial automation, commercial refrigeration compressors, and process control equipment.

Not electrical technology. Not manufacturing reshoring. Not Industry 4.0.
Just Emerson, at the moment of maximum uncertainty — 1890 — placing every chip on the AC electric motor. And then gradually installing that motor into the world's compressors, valves, fans, and automated control systems.

Emerson was born in 1826 to a farming family in New York State. He became a judge in Illinois — but his personal passion was invention. In 1890 he witnessed the spectacular scene of Westinghouse lighting up the entire Chicago World's Fair with alternating current — then he sold off most of his non-electrical assets and founded the Emerson Electric Manufacturing Company. Its first product was a small single-phase AC induction motor — the ancestor of the motor found in every home appliance today.

But Emerson's true genius wasn't invention — it was "using acquisitions to buy technology and improve its management." In 1954, CEO Wallace Parsons launched what has been called "the most sustained and effective diversification through M&A in the history of management — over 200 acquisitions, almost no major failures." Emerson didn't buy turnaround companies — it bought mid-sized industrial companies that "owned proprietary technology but were delivering low profits under bad management." After buying, Emerson implanted its "Best Cost Management" methodology — a set of meticulous budgeting and production efficiency assessment disciplines known internally as the "Profit Plan" — and drove profits up.

Chuck Knight (not the architect — Charles Knight, CEO from 1973 to 2000) drove Emerson's profits to grow for 43 consecutive years — earning it the reputation as "the best-managed industrial company in the world." Upon retirement, he said: "We never invented a single world-changing technology. We just knew better than the people who invented those technologies how to price, how to produce, and how to manage inventory."

Three things.
First: bet on a law of physics — not on a company.
Emerson wasn't betting on whether Westinghouse or Edison would win — he was betting that the laws of AC were better suited to long-distance transmission than DC. Laws don't go bankrupt — companies can.
What you're betting on — is it a specific company's success or the irreversibility of a physical or technological law?
Second: don't buy turnaround companies — buy mid-sized companies that "have proprietary technology but problematic management."
Emerson grew through 200+ acquisitions, but it never bought companies that needed "rescuing." It bought companies where "management wasn't trying hard enough" — switched in better management, and the profits emerged on their own.
Your acquisition criterion — is it "technology scarcity" or "lazy management"?
Third: management itself can be made into a replicable product.
Emerson's "Profit Plan" is not a philosophy — it's an Excel-level tracking system, granular down to every workstation, every compressor, every customer contract, with cost ratios and target profit margins.
Is your management — a replicable system or your improvisational intuition?

In his autobiography, Charles Knight wrote: "In my twenty-seven years as CEO, I made decisions on my own only five times — everything else was automatically driven out by the Profit Plan system. I managed Emerson — Emerson was not managed by me."

← Previous Contents Next →
Chapter 156 Regeneron

1988, Tarrytown, New York. A young neuroscientist named Leonard Schleifer — who had spent years studying nerve growth factor at Cornell University Medical Center — shut himself inside a rented lab in Westchester County. Together with his chief scientist George Yancopoulos (a Columbia University immunology postdoc), he scraped together $1 million in angel funding and set out to do one thing with mouse genes: make a mouse grow a complete human immune system, identical to a human's. This is not a metaphor — insert the entire human antibody gene repertoire into the mouse genome via transgenic technology, so that an immunodeficient mouse acquires a fully human B-cell immune response. Today Regeneron is worth roughly $100 billion.

Not CRISPR. Not pharma. Not government R&D.
Just using humanized mice to solve a nightmare that had tortured the pharmaceutical industry for decades: how to obtain high-affinity, fully human monoclonal antibodies against human targets — without conducting human experiments.

Before Regeneron, monoclonal antibody drug discovery was maddeningly cumbersome: immunize a mouse to get a mouse antibody, then use molecular biology techniques to "graft" the mouse antibody's variable region onto a human antibody framework — a process called "humanization." Every step was agonizingly slow, astronomically expensive, had a low success rate, and could trigger a human anti-mouse antibody response.

The Schleifer-Yancopoulos solution was to use homologous recombination technology in embryonic stem cells to knock out all of the mouse's antibody genes (the full V-D-J repertoire) and replace them with the corresponding human genes. The result: when you inoculate these mice with a human target, they produce genuinely human antibodies — no humanization needed, no engineering required. This platform was named VelociMouse → the corresponding antibody screening process was called VelocImmune → and the corresponding ultra-rapid antibody gene sequencing was called VelociGene.

Aflibercept (Eylea) — an intravitreal injection for treating age-related macular degeneration, approved by the FDA in 2007 — was Regeneron's first mega-blockbuster. It's a fusion protein that combines a VEGF receptor fragment with the Fc tail of IgG, binding all VEGF signaling molecules with affinity up to a hundred times higher, essentially "licking them all clean" and preventing abnormal blood vessel leakage in the eye. One shot — annual sales once exceeded $8 billion.

Then came Dupixent (dupilumab) — not wholly made by VelocImmune, but co-developed by Regeneron and Sanofi — approved in 2017 for atopic dermatitis and later expanded to asthma, eosinophilic esophagitis, and prurigo nodularis — all "type 2 inflammation"-related diseases blocked at the IL-4Rα antibody target. Dupixent currently generates over $10 billion in annual sales, making it one of the best-selling anti-inflammatory biologics on earth.

The 2020 COVID pandemic — Regeneron produced REGEN-COV (the casirivimab + imdevimab cocktail antibody therapy), which was part of President Trump's treatment regimen at the time. Although it was soon overtaken by viral variants, it played an important role in prevention and treatment during the 2020–2021 window before vaccines. In that pandemic, Regeneron demonstrated the full chain of speed: "rapid antibody discovery → manufacturing → clinical trials."

Three things.
First: put an entire human immune system into a mouse — and then let the mouse invent human antibodies for you.
This is "immune system simulation" in a physical sense. The human antibodies produced by VelocImmune mice — are not computer-designed, not chemically synthesized — they are the product of immune evolution, naturally selected inside a mouse's spleen.
Can you "replace an algorithm with a physical simulation" — rather than "replace physical simulation with an algorithm"?
Second: antibody affinity determines whether you're a $1 billion drug or a $10 billion drug.
Aflibercept's binding affinity for VEGF is more than 100 times that of its natural receptor. That affinity isn't "a little higher" — it's at the level of "clearing out every last signaling molecule." That hundredfold gap is the gap in drug efficacy.
Is your product — "good enough" or "physically saturated"?
Third: a single biological pathway — can underpin ten entirely unrelated diseases.
Dupixent's target, IL-4Rα, sits in the type 2 inflammation pathway. That pathway governs atopic dermatitis, and also governs asthma, and also governs nasal polyps, and also governs eosinophilic esophagitis. One switch — six diseases. Six — with one drug.
Can your product — unlock a pathway — and then translate sideways across indications?

Leonard Schleifer remains CEO to this day. In November 2020, before the U.S. market opened, he said on Bloomberg TV: "We don't study viruses. We study — how to use an antibody to completely block any target. Change the target — same platform."

← Previous Contents Next →
Chapter 157 Apollo-Global-Management

1989, New York. Three young men who had come out of Drexel Burnham Lambert — Michael Milken's junk bond empire — Leon Black, Josh Harris, and Marc Rowan — rented a small office to do a kind of business that "basically no one dared to do upfront": using junk bond financing — to buy companies. The first person who called was Republic Steel. The business was called "leveraged buyouts" — and the biggest players at the time were KKR and Blackstone. Black founded Apollo — named for the ancient Greek sun god — because "the sun shines on companies that no one else wants to look at." Today Apollo Global Management manages over $600 billion in assets.

Not private equity. Not buyouts. Not DCF models. Not junk bonds.
Just an extraordinarily specific investment philosophy: "Buy the companies no one wants to touch — and then fix them."

Leon Black was born in 1951 to a Jewish family in New York. His father was the CEO of United Brands — and in 1975 he jumped to his death after an illegal political contributions scandal. Black was 24 years old at the time. He later went to Drexel Burnham Lambert — doing M&A financing under Michael Milken, participating in some of the biggest deals of the 1980s junk bond wave. Drexel went bankrupt in 1990 from a junk bond scandal — and Black picked up the team he'd brought with him from the rubble and founded Apollo.

Apollo's first major bet was buying Executive Life Insurance Company in the early 1990s — a junk-bond-backed insurer — acquired at a government auction during its bankruptcy liquidation. This deal was eventually investigated, prosecuted, and fined — and Black himself bore a prolonged reputational cost for it. But Apollo extracted enormous value from Executive Life's residual assets — and rolled those assets into the seed pool for several larger funds that followed.

Apollo's style is starkly different from Blackstone and KKR — it specializes in "distressed assets," bankruptcy restructurings, and acquisitions of troubled insurers, banks, and industrial companies. Its credo: "We go shopping where other people call the police."

The 2008 financial crisis — Apollo used its distressed-asset funds to buy up massive quantities of subordinated bonds and bankrupt, restructured financial assets at deep discounts. Then in the 2010s it expanded from "distressed-asset specialist" into insurance annuities — founding Athene Holding, integrating the fixed-income assets of pensions and retirement annuities under Apollo's investment platform. Athene is now both a client of Apollo (buying distressed loans and structured credit that Apollo manages) and one of the largest sources of Apollo's assets under management.

Three things.
First: bet on places where other people call the police — but you'd better truly know how to fix things.
Apollo doesn't "buy low and wait for it to go up" — it buys bankrupt companies, replaces management, restructures the capital stack, and sometimes holds for seven or eight years until the fix is complete. Internally it has a unit called the "Apollo Operations Group" — dedicated to embedding people inside acquired companies to execute operational improvements.
Do you have the capability to "fix" a company or an asset — rather than just "pick up a bargain"?
Second: insurance float — play it the way Buffett does.
Apollo bought Athene — an insurance company that sells annuity products. An insurer collects annuity premiums — and must pay out fixed future retirement income — and the money in between can be invested. Apollo uses this float to buy credit funds and structured assets that it created itself — dead money (premiums) in the left hand, high-yield debt in the right hand.
Where is your source of "low-cost, long-duration capital"?
Third: don't think about being first — think about "who wants this asset the most?"
In internal meetings, Black repeatedly told his partners: "We don't bid the highest price in an auction. We go after the things — that only we have the guts and the balance sheet to dare to buy."
Do you win — because you have more money than your competitors, or because you can endure a longer repair cycle than they can?

Leon Black stepped down as CEO in 2021 for personal reasons — succeeded by Marc Rowan. Rowan still says the same thing in every orientation meeting with a new fund manager: "Every time you open the Wall Street Journal — find the page where the headline reads 'This company cannot be fixed.' That's Apollo's first phone call."

← Previous Contents Next →
Chapter 158 Marathon-Petroleum

1887, northwestern Ohio — near the village of Findlay. A small crude oil producer named John Vanderlice poked the region's first commercial oil well in the middle of a cornfield. When the oil spouted up, he had no pipeline, no refinery, no storage tanks. He loaded wooden barrels onto a farmer's wagon — hauled the crude by horse to the nearest railroad siding, and every Wednesday morning loaded it onto a train headed for Standard Oil's refineries. Standard Oil had by then been consolidated by John D. Rockefeller into a monopoly machine. But soon enough, Rockefeller's Standard Oil was forcibly broken up by the Supreme Court in 1911 — and one piece of it later became Marathon. Today Marathon Petroleum is the largest independent refiner in the United States, worth roughly $70 billion.

Not oil. Not shale. Not gas stations. Not crude prices.
Just "cracking" — the physical process of distilling a barrel of crude oil into gasoline, diesel, jet fuel, and petrochemical feedstocks. Marathon is a specialist player in the "refining" middle link — it doesn't produce crude, and it doesn't dominate end-user retail. What it earns is the crack spread: the spread between the price of crude oil and the price of refined products.

1888 — when the Ohio Oil Company (Marathon's predecessor) was founded, it had four employees, one well producing 20 barrels a day, and a small atmospheric distillation column. It was quickly bought by Standard Oil as one link in the pipeline network. In 1911 the Supreme Court ordered the breakup of Standard Oil — Ohio Oil was "liberated" and regained its independent identity. It then began laying down its own pipeline and refinery network across the American Midwest — Illinois, Indiana, Ohio, Michigan.

The 1930s — Ohio Oil acquired a small Texas refining company named Marathon. They liked the name — and in 1962 they renamed the entire company Marathon Oil Company. Marathon supplied gasoline and aviation fuel for American car culture and American warships and aircraft during World War II — then went through its first enormous windfall during the 1970s oil crisis because it owned the Cook Inlet fields in Alaska.

2011 — Marathon Oil split itself in two: upstream (oil exploration and production) = Marathon Oil; downstream (refining + pipelines) = Marathon Petroleum. After that, Marathon Petroleum executed what may have been the most important acquisition in downstream oil history — buying Andeavor (formerly Tesoro) for $23 billion in 2018 — making itself the largest refiner in America overnight, with crude processing capacity exceeding three million barrels per day.

Three things.
First: refining — it's just "being a middleman" — but be the most extreme middleman there is.
Marathon doesn't bet on the direction of oil prices. It's not an exploration company — it buys crude, processes it, and sells the output. What it captures is that crack spread. What determines whether it thrives isn't the absolute level of oil prices — it's the spread between "crude price / refined product price" and the conversion rate determined by the complexity of its refineries.
Does your profit — come from "raw material prices" or from "the processing spread"?
Second: pipelines and logistics — these are the true moat around refining profits.
Marathon's pipeline subsidiary MPLX controls tens of thousands of miles of pipelines and terminal facilities. These pipelines don't make money on the price of oil — they make money on a per-barrel transportation fee. When oil drops from $100 to $40, refinery profits can go to zero — but the pipeline fees keep coming in.
Do you have a "fee income" stream like a pipeline — where regardless of whether your customers are making money or losing it — you still collect that "toll"?
Third: buy the massive old rival — and then integrate the logistics.
When Marathon bought Andeavor, Wall Street thought it was buying at the top. But Marathon wasn't "buying the dip" — it was buying the "system": Andeavor's network of refineries + pipelines + gas stations on the U.S. West Coast and Midwest — complementary to Marathon's own Midwest and Gulf Coast systems. After the merger — backend logistics integration eliminated the cost of "competing against yourself" for crude transport and pipeline capacity on overlapping routes.

John D. Rockefeller died in 1937 — he never once acknowledged having swallowed and operated the "Ohio Oil Company" back in the 1880s. But Marathon Petroleum's headquarters building still sits in Findlay, Ohio — in the exact same spot where that first well was drilled next to a cornfield.

← Previous Contents Next →
Chapter 159 3M

1902, Two Harbors, Minnesota. Five local businessmen — a doctor, a lawyer, two grocers, and a railroad ticket agent — gathered together and signed a partnership agreement. They had bought the mining rights to a corundum deposit on the north shore of Lake Superior — they believed crushed corundum could serve as the abrasive for sandpaper and grinding wheels. On the first day of mining — the stone was blasted out. Corundum? There was none. It was anorthosite — a low-grade mineral whose hardness sits somewhere near chalk on the hardness scale. The "mine" in their hands — was in fact a worthless pile of waste rock. Today 3M is worth roughly $70 billion, with products in over 200 countries. Post-it Notes, Scotch masking tape, N95 respirators — these are just the tip of the iceberg.

Not materials science. Not industrial adhesives. Not abrasive technology.
Just "making sandpaper out of the wrong rock — and then being forced to invent the right thing out of the wrong one."

In 1902 the five men paid $5,000 for the mine — and discovered not only that they had no corundum, but that they had no money. They didn't shutter the company — instead they started synthesizing their own abrasives from scratch in another small Minnesota town. Failure after failure — the early abrasive grains wouldn't stay on the paper; they'd go on, then come off as a sticky mush. In 1905 they finally produced one barely-usable sheet of corundum sandpaper — the company was days away from its third bankruptcy.

Then 3M grew out of sandpaper two fiber-based technology streams that would later prove profoundly important to the entire world: first, coating — the process of applying something uniformly onto a substrate (first abrasives, then adhesives, then magnetic powder, then pharmaceuticals, nanomaterials, and optical films); second, adhesives — capable of bonding to any surface, removable, and weather-resistant enough to stay on an airplane wing for thirty years without cracking or peeling.

1925 — a lab technician named Richard Drew invented Scotch masking tape. The story is simple: he went to an auto body shop to interview sandpaper customers — and watched painters use newspaper and glue to mask the dividing line on two-tone cars, always pulling off a strip of freshly sprayed paint when they tore it away. Drew spent two years testing hundreds of adhesive formulations — and finally produced "tape you can peel off without pulling the paint with it." Then, following the same logic, he invented Scotch transparent tape — during the 1930s Depression, when people saved money by not buying new things and instead used tape to mend old books, broken toys, and cracked windowpanes.

1968 — Spencer Silver accidentally synthesized a weak adhesive in a 3M lab — one that "would stick but could be peeled off and re-stuck easily." At the time he considered it a failure. Then his colleague — a chemist named Art Fry — found, while singing in his church choir, that the little pieces of paper he used as bookmarks kept falling out. He thought of Silver's weak adhesive, the kind that wouldn't damage paper. In 1977 they issued the first internal trial batch of "Press 'n Peel" — later renamed the "Post-it Note" — and launched it nationwide in 1980.

Three things.
First: take the wrong rock — then start synthesizing your own sandpaper.
The rock in 3M's mine was worthless. The company was forced to either shut down — or synthesize its own chemical abrasives. That "being forced" formed its DNA — 3M has never been the kind of company that "owns scarce resources." It is the kind of company that "uses existing chemical and physical knowledge to synthesize new materials itself."
Is your core — "owning good raw materials" or "being able to make raw materials"?
Second: 15% free time — let materials scientists stumble into serendipity in their own experiments.
3M's famous "15% time" culture allows technical staff to leave assigned projects for a stretch and do their own experiments. The Post-it's weak adhesive — was a failure within the assigned project. Within that 15% time, it collided with the right use case.
Do you give your core technical staff a period of "no assigned goal" free experimentation time?
Third: a company, after more than a century, still lines up on a few of the most fundamental physical properties — stick, coat, grind, filter.
3M doesn't do "strategic pivots" — it does "taking the same coating platform and translating it sideways onto new substrates." From sandpaper → tape → reflective film → LCD brightness-enhancement film → N95 filter media — every one of them is coating a thin film, bonding it to a different substrate.
Can your foundational process platform — be translated sideways into ten completely unrelated downstream applications?

The last of the five founders — Henry Bryan — left the company in 1918. In his later years he said: "If I had known that rock was anorthosite — I'd still be selling groceries today. Ignorance is a three-letter blessing."

← Previous Contents Next →
Chapter 160 Valero-Energy

1980, San Antonio, Texas. The CEO of a pipeline and natural gas company called Coastal States Gas Corporation — Oscar Wyatt — spun off his downstream refining assets to shareholders, forming a new company named Valero. The name was taken from an old Spanish mission from San Antonio's earliest days: Mission San Antonio de Valero — otherwise known as the Alamo. Wyatt said the name meant: "We can face any numerical enemy alone." Today Valero Energy is one of the largest independent refiners in the world, worth roughly $50 billion.

Not oil. Not pipelines. Not ethanol. Not biodiesel.
Just a business logic stripped down to its barest essentials: build refining capacity as close as possible to the crude-producing regions — but as far as possible from every other refinery — and then earn the crack spread through logistics exclusivity.

Wyatt was born in 1925 in rural Texas — no college degree. He started by collecting rents on natural gas pipelines, then gradually acquired small and mid-sized pipelines — and stitched them into a regional network using self-built connectors. Coastal States was the largest independent natural gas pipeline operator and downstream refiner in the U.S. Southwest during the 1960s and 70s. But in 1980 Wyatt decided to split — because refining profits and natural gas pipeline profits operate on two entirely different oil-price cycles. After the split, Valero focused solely on refining.

Valero's early refineries were all situated in southern Texas — close to the Gulf of Mexico's crude import terminals and the domestic pipeline hub at Corpus Christi. The geographic logic was this: refineries in the U.S. Midwest and East Coast have to transport crude in by pipeline or tanker from thousands of miles away; Valero's refineries receive crude directly from tankers at Gulf import terminals — distances measured in yards, not miles.

After 2000, Valero made two decisive lateral expansions: first, acquiring ethanol plants — putting corn ethanol and conventional refining onto the same fuel distribution network (because America's Renewable Fuel Standard requires gasoline to be blended with more than 10% ethanol); second, acquiring overseas refineries in the U.K. and Canada — like Pembroke and Milford Haven — deploying the same "close to port" logic across both shores of the Atlantic.

Valero is not a brand. Valero has no gas stations. Valero doesn't sell anything that comes in a package. It sells one kind of intermediate product — the gasoline, diesel, jet fuel, asphalt, and sulfur that flow out of its refineries — poured via pipeline, rail, and tanker into the distribution networks of the multinational oil companies (Shell, BP, ExxonMobil). Valero is the "raw gasoline supplier."

Three things.
First: location, location, location — but the location that matters is "how far from the international crude import terminal."
Crude oil is one of the heaviest and least expensive logistics commodities in the world — the cost of shipping a single barrel, as a percentage of total cost, can get large enough to swallow the entire crack spread. Valero nailed its refineries right next to the docks — which is effectively an automatic percentage-point boost to its crack spread, courtesy of near-zero transportation cost.
Is the distance between your factory and its "raw material source" — part of your pricing?
Second: don't own gas stations — be "the Intel inside the gasoline."
The gasoline inside every car's tank may have been bought at a Chevron station — but the gasoline itself was very likely produced at a Valero refinery. Gas stations aren't Valero's business — it's the "fuel module" supplier.
Is it possible — for you to not be a consumer-facing brand — and instead just be the "core component inside someone else's brand"?
Third: regulations force demand for ethanol — so Valero just buys ethanol plants.
The U.S. government mandated ethanol blending — and Valero didn't want to depend on ethanol suppliers. It directly bought over a dozen ethanol plants — converting a regulation from an "external risk" into "internally managed procurement."
Do you have a way to turn a "regulation" into a link in your own internal value chain — rather than being a passive compliance actor?

Oscar Wyatt died in 2011. During the last decade of his life, for the most part he stayed out of Valero's operations — instead spending his time funding the construction of several small wildlife refuges and bird habitats in southern Texas. He said: "Where gasoline can't move — let the migratory birds fly over the Alamo."

← Previous Contents Next →
Chapter 161 Illinois-Tool-Works

1912, Chicago. Byron C. Smith — a German-American machinist — opened his first workshop in the industrial district near the Union Stockyards. The shop had only one business: making a better, more precise, less failure-prone gear drive system and bed slide assembly — upgrading the aging machine tools of Chicago's rapidly expanding meatpacking, printing, and metal-stamping factories. Today Illinois Tool Works (ITW) is worth roughly $80 billion in market cap, with over 80 business divisions worldwide that manufacture automotive welding consumables, construction fasteners, food packaging equipment, test and measurement instruments, and polymer adhesive components — a "general store" of industrial goods, yet every single niche category ranks among the top two globally.

Not focus. Not core technology. Not a patent moat. Not a single product category.
It was the "80/20 management method" — forcing every small division's operating model, per the Pareto Principle, down to the most critical 20% of products and customers.

Byron Smith was born in Chicago in 1887. He had discovered a pervasive problem that tormented every small and mid-sized manufacturer: the drive gears and sliding beds you could buy — from the very same machine tool factory — varied wildly in noise, vibration, and precision. Because nobody had inspected them under a standardized tolerance process before they left the factory. Smith said: "What I want to make isn't a better gear — it's a quality inspection method that makes the thousandth gear every bit as good as the first."

And then ITW, over the next hundred-plus years, did not choose to "be a focused machinery company." It chose what business schools call "one of the most extreme distributed management models in existence": acquiring hundreds of small, leading manufacturing enterprises, turning each into an independent mini-division, and then implanting ITW's three-part toolkit into every division — 80/20 profitability analysis, "Inside-Out" strategic analysis, and "Customer-Back Innovation."

80/20 doesn't mean "only big customers matter." It means: cut the unprofitable SKUs from 80% of the product line, cut the customers from 80% of the roster who place one order and vanish, shrink the management layers until each division has no more than three types of people — finance, engineering, and sales. Then that division will automatically take its small, profitable niche and become number one or number two in its segment.

ITW's product line is an "industrial general store" — from welding wire for automotive body shops, to plastic cable ties for tying rebar, to food packaging machines for fast-food restaurants — utterly unrelated to one another. But behind this seemingly chaotic assortment lies this: it only acquires consumable businesses with "repeat consumption rates." The welding wire on an auto plant floor runs out every day and needs restocking. The plastic cable ties on a construction site get used by the thousands on every project. The restaurant's packaging equipment needs film rolls replaced every single day. Every ITW product — is an industrial consumable that "gets swapped every shift, every day."

Three things.
First thing: the management method — 80/20 cuts products, cuts customers, cuts management layers — is more valuable than any industrial technology.
ITW's core IP isn't some patent. It's the 80/20 methodology training that teaches division managers to kill unprofitable SKUs.
Your management — can it be distilled into "a formula" and reused across different industries?
Second thing: the industrial general store — every single item is low-value but consumed daily without fail.
Welding consumables, fasteners, packaging film, measurement probes, adhesives — unit prices range from pennies to a few dozen dollars — but they get consumed every day, every shift, every stamping cycle, every packaging run, every die change.
Your product — is it "purchased once and done" or "consumed through the process of use"?
Third thing: acquisitions — but immediately followed by "division miniaturization."
ITW buys a company, immediately guts the headquarters, collapses the original six management tiers down to two or three key people, and slashes the product catalog to only the 20% of SKUs that make money. This "hara-kiri integration" makes every acquired asset deliver an ROIC several times higher than before, over the long run.

After Byron Smith retired in the 1960s, the board followed the succession rules he himself had laid down — and donated all his wealth to employees and the foundation. He said: "ITW is not a company. It's a formula. Once you learn the formula — you don't need me anymore."

← Previous Contents Next →
Chapter 162 Kinder-Morgan

1997, Houston. Richard Kinder — who had spent twenty years as COO and Vice Chairman at Enron — had left Enron a year earlier after falling out with CEO Ken Lay. He began doing something he had pushed for relentlessly at Enron but Lay had shot down: using low leverage to acquire dormant assets — pipelines. Kinder and his college classmate Bill Morgan had jointly run Enron's pipeline subsidiary, Enron Liquids Pipeline, during their Enron years. They used $40 million of their own money (from selling their Enron stock) as seed capital and bought their first natural gas liquids pipeline from a bankrupt conglomerate. Today Kinder Morgan is North America's largest independent pipeline infrastructure company, operating nearly 83,000 miles of pipelines — with a market cap of roughly $70 billion.

Not shale gas. Not energy independence. Not LNG terminals.
It was "collecting a few cents in tolls on every gallon of natural gas and gasoline that flows through the pipes."

Kinder was born in Missouri in 1944. He earned dual degrees in engineering and law from the University of Missouri — then joined the military as a JAG prosecutor. After leaving the service he worked as legal counsel at a natural gas pipeline company in Florida, gradually rose into operations management, and was then recruited into Enron's senior leadership. During his time at Enron he ran the pipeline business — dull, steady "infrastructure" income — but it was eclipsed in Enron's culture, because Wall Street was chasing Ken Lay and Jeff Skilling's "energy trading revolution" and "digital broadband network."

Kinder's conflict with Lay was fundamental: Lay wanted to redeploy the capital from Enron's pipeline business into broadband and power trading. Kinder said the pipelines were Enron's only stable cash anchor — mortgaging them to back trading bets was dangerous. Less than five years after Lay pushed him out — Enron collapsed under off-balance-sheet liabilities and hidden derivative losses. The best pieces of Enron's post-bankruptcy pipeline assets — were bought up by Kinder Morgan's founders for a few hundred million dollars.

Kinder Morgan is not a pipeline builder — it's a pipeline operator. Its business model is collecting transportation fees. Whether natural gas is $2 or $10, every million BTU traveling a thousand miles through its pipeline network pays a set rate. Most of these fees come from long-term contracts or federally regulated "cost-plus-return" rate structures — meaning Kinder Morgan's cash flows are highly stable, resembling those of a public utility.

In 2013 — weighed down by ever-increasing distribution obligations under its master limited partnership structure — Kinder Morgan consolidated four publicly listed entities into a single C-corporation. After the merger the stock price first dropped, then rose — but Kinder took a symbolic salary of $1 a year. His entire income came from dividends.

Three things.
First thing: don't bet on commodity prices — bet on throughput.
Kinder Morgan's pipelines don't care whether natural gas is $2 or $10 — they only care that more gas needs moving. The American shale revolution sent throughput volumes soaring — whether or not the upstream companies made money, Kinder collected the tolls.
Is your revenue model — sensitive to "price" or to "volume"?
Second thing: you were right — and your former company will blow up.
Kinder had warned as early as 1997 that Enron should not mortgage its pipeline assets. By 2001 Enron went bankrupt — and then Enron's most valuable downstream pipelines were bought by Kinder at rock-bottom prices.
Have you spotted a mistake your former employer is making — one that will kill it years from now — while you stand by, ready to pick up the wreckage?
Third thing: a CEO takes a $1 salary — a lifetime's wealth comes entirely from dividends and the capital appreciation of his own stock.
Kinder was willing to let the stock price languish in the market's cold shoulder for long stretches — because he knew the dividend yield would eventually pull the share price back. He was playing a multi-decade game of dividend compounding, not quarterly earnings surprises.

Richard Kinder lives in Houston now. On the wall at the entrance to his office building, there is a line etched in bronze: "We don't trade energy — we move it."

← Previous Contents Next →
Chapter 163 Hilton

1919, Cisco, Texas. A young man named Conrad Hilton — he had originally intended to buy a small bank in Cisco. The negotiations fell through. Dejected, he walked into a little hotel across from the bank called the Mobley Hotel — it was crammed with oil workers, so packed that beds were rented in eight-hour shifts. He bought the hotel that very night — using the last of the money in his pocket for the down payment. Then he crammed beds into every corner of the lobby. Then he went on to buy every hotel he could find in every oil town across Texas. Today Hilton Worldwide is one of the largest premium hotel groups on the planet, with 7,600 hotels and 1.2 million rooms — worth roughly $70 billion in market cap.

Not tourism. Not aviation. Not business travel. Not real estate.
It was failing to buy a bank — walking past a hotel — discovering that "rotating beds" yielded ten times the interest of a fixed deposit — and then spending a lifetime buying and operating hotel properties.

Conrad Hilton was born in 1887 in a small trading post in New Mexico — his father was a Norwegian immigrant who ran a general store. After finishing his studies at the New Mexico Military Institute, he helped keep the books at the family grocery. During World War I he served in the military. After being discharged, he set out with a total of $5,000 to break into banking — but every little bank in Texas was asking exorbitant prices. The Mobley Hotel accident was the greatest serendipity of his life — he discovered that the hotel charged for beds in eight-hour blocks — three shifts a day — the same room could be sold three times over.

Through the 1920s, Hilton bought hotels in Dallas, Fort Worth, and El Paso using one mortgage after another. During this period he completed his flagship of the moment — the Dallas Hilton — just days before the stock market crash of 1929 — and then slammed headlong into the Great Depression. Hotel occupancy rates plunged below 10%. Hilton was hounded by creditors, had his water cut off, had his linen supply halted — but he did not go bankrupt. In every hotel restaurant, he added cheap set meals to draw locals in to dine — using the meager income from "neighborhood restaurants" to prop up the hotels' operating costs.

In the 1940s and 50s he placed an even bigger bet — from New York to Chicago to Puerto Rico — buying up America's most iconic Old World luxury hotels one after another: the Waldorf Astoria (New York), the Stevens Hotel (Chicago — the world's largest hotel, later renamed the Chicago Hilton), and the Plaza Hotel (New York). In 1949 Hilton became the first company to operate an American hotel chain outside the continental U.S. — opening the Caribe Hilton in San Juan.

Three things.
First thing: see the value of "reuse" in a place nobody else thought to look.
What Conrad saw at the Mobley Hotel was "the same room being bought by three shifts of oil workers in a single day" — this is the physical origin of what the hotel industry would later call "RevPAR" (Revenue Per Available Room). In 1919 he was already doing the math in his head: "maximum sellable instances per room, per day."
Your asset — in a single unit of time, how many different customers can it be sold to?
Second thing: in a depression — a hotel does not survive on overnight guests. It survives on the local restaurant.
Hilton's Depression-era strategy was profoundly contrarian — he didn't just wait for travelers to return. He turned every hotel restaurant into a discount lunch spot that drew local residents. He used a non-core business to transfuse the core business — and survived the lethal trough of 1930 to 1933.
When the economy is terrible — what "local restaurant" replaces the "overnight guests" you lost?
Third thing: buy the oldest, the most iconic — and then preserve their facades, their names, and their souls.
Hilton bought the Waldorf — and kept the name "Waldorf Astoria." Bought the Stevens — and kept the name "Conrad Hilton." He didn't carve his own surname onto old legends — he let the legends carry his management standards.
When you acquire — is it "rename and rebrand" or "preserve the soul"?

Conrad Hilton died in 1979. His will bequeathed his entire fortune to the Conrad N. Hilton Foundation — devoted primarily to education for nuns and humanitarian projects around the world. He had a habit of placing a Bible in the lobby of every newly opened hotel. He defined the hotel business as "turning the foreigners of the earth — into neighbors for a night."

← Previous Contents Next →
Chapter 164 Coherent

1966, Palo Alto, California. A group of young men out of Stanford's electrical engineering and physics departments — led by Eugene Watson — rented a small workshop under 500 square feet in the Stanford Industrial Park and began manufacturing industrial carbon dioxide (CO₂) lasers. Before that, lasers were almost exclusively laboratory apparatuses — mirrors bolted onto optical tables, scientists manually turning screws to align the beam. Watson's idea was: take the CO₂ laser off the lab bench and onto the factory floor — to cut metal, weld plastic, engrave wood. Later, the company would come to be called Coherent. Today Coherent is worth roughly $30 billion in market cap.

Not AR/VR. Not 3D sensing. Not autonomous driving.
It was the least sexy idea of 1966 — "turning the world's most precise physics instrument into a tool you switch on with a foot pedal on a factory assembly line."

Eugene Watson worked a few years at Spectra-Physics before striking out on his own — he had discovered that the application potential of CO₂ lasers was bottlenecked by two things: first, output power was wildly unstable — a perfectly tuned power level would drift 3% by afternoon; second, the optics were extremely sensitive to dust — cleanroom operation was mandatory. And factory floors were full of oil, dust, and vibration.

In 1969 Watson launched the first "industrial sealed CO₂ laser" — the sealed-tube design kept workshop dust from ever reaching the resonance cavity. All the operator had to do was push a button. Then lasers exploded across acrylic signage, automotive dashboard cutting, and textile fabric cutting.

After the turn of the 21st century, Coherent acquired and consolidated several laser competitors including Rofin-Sinar, and in 2022 completed a merger with II-VI Incorporated — a giant in infrared and semiconductor laser components. The merged Coherent maps across nearly the full commercial laser wavelength spectrum, from ultraviolet to far-infrared — delivering vertically integrated capability from laser crystals, pump diodes, and optical fiber all the way to complete laser systems.

Three things.
First thing: turning a lab instrument into a factory tool — through a single physical design, the sealed tube.
The barrier between lasers and factory-floor applications, from scientific experiment to industrial deployment, wasn't "power isn't high enough" — it was "not rugged enough." Watson solved it with a single sealed tube.
Your product — can a single "sealed tube" kind of physical modification expand its environmental applicability tenfold?
Second thing: vertical integration — from II-VI's crystals to pumps to fiber to systems.
After acquiring II-VI, Coherent gained direct control over the most expensive raw materials inside a laser (semiconductor laser pump diodes, infrared and nonlinear optical crystals) — meaning competitors could buy its light-emitting components but could never go below its internal transfer price.
Can you use "acquiring your suppliers" to ensure competitors can never be cheaper than you?
Third thing: the name is "Coherent" — in physics this means "in phase" — when a beam of light is in phase, it is a laser.
The company's name itself describes a physical concept. Watson said: "If your customers don't know your principles — let them read your name."

Eugene Watson retired in 1994 and remained in the Bay Area. In his later years he invested in high-school science labs around the Bay — and on the wall of every lab hung a single question: "Did you adjust a parameter today and write it down?"

← Previous Contents Next →
Chapter 165 American-Electric-Power

1906, Columbus, Ohio. Richard E. Breed — an electrical engineer out of the Chicago Edison Company — consolidated dozens of small, early-generation power plants and electric companies scattered across Ohio's villages and towns into a holding entity called American Gas and Electric Company, then gradually renamed it American Electric Power (AEP). At the time, there was no unified national grid in America — every town generated its own power, with different voltages, different frequencies, some using AC, some still on DC. Breed's consolidation logic was exceedingly simple: you link these dozens of small grids into a synchronized AC network — a single long line lets one town's surplus power be used in another town, and power plant capacity utilization jumps overnight. Today AEP is worth roughly $50 billion in market cap and is one of the largest electric generation and transmission utility companies in the United States.

Not electricity market liberalization. Not clean energy. Not nuclear fusion.
It was forcibly welding dozens of small generating stations — which couldn't even agree on a frequency — together with transformers and high-voltage transmission lines.

Breed was born in Illinois in 1865 and rose from a rank-and-file technician to grid dispatch engineer at the Chicago Edison Company. He saw that the AC transformer's ability to "convert frequency and transform voltage" could be used to stitch incompatible local mini-grids into a nationwide network — just as the "standard gauge" of the railroad companies had once patched together all the short-line railroads into a continental rail network.

AEP's core asset was never power plants — it was "high-voltage transmission lines." By the late 1950s, AEP's engineers had constructed America's first 345 kV extra-high-voltage transmission corridor in western Virginia — enabling electricity to travel economically from West Virginia coal plants to cities hundreds of miles away, as far as Columbus and Fort Wayne. In the 1970s it further developed 765 kV transmission technology — still the highest transmission voltage in North America today — allowing a single power plant to serve the needs of several states.

In the 21st century, AEP became one of the companies facing the hardest "coal transition" — because policies across the Midwest and East Coast, along with EPA rules, forced it to shut down coal plants and switch to natural gas and renewables, all while maintaining grid stability and keeping electricity rates in check. It is now investing heavily in interstate transmission lines — to move wind and solar power from the Western Plains to the population centers of the East.

Three things.
First thing: the grid's value is not in "where the electricity comes from" — it's in "whether the electricity can be delivered to wherever it's needed."
Breed didn't build power stations — he built connections. Voltage times distance divided by losses — that's the fundamental equation of grid engineering. He built AEP's moat on "interstate transmission approval rights" and "physical possession of critical transmission corridors."
In your industry — is the core bottleneck "production" or "transmission"?
Second thing: standard gauge is worth more than locomotives.
AEP unified frequency, unified voltage, unified dispatch — in essence, it physically imposed a "standard gauge" for electricity across the American Midwest.
In your industry — is there a "fragmented interface" waiting for you to unify it?
Third thing: from coal to wind — the grid company's core competency is "conversion."
When coal plants are shut down — AEP has to integrate wind and solar into the same network. But wind and solar are not in the same locations as coal mines — so new transmission corridors are needed. AEP's "new growth" is transmission infrastructure — not generation.

Richard Breed died in 1945. His successors, to this day, are still building extensions of the first high-voltage transmission corridors he sketched on paper in the 1910s. The tower sites haven't changed — just taller towers, thicker conductors.

← Previous Contents Next →
Chapter 166 EOG-Resources

1999, Houston. Enron Oil & Gas — known as EOG — saw its parent company Enron, under instructions from Ken Lay, decide to spin off this "boring" oil exploration subsidiary and list it independently, so the capital freed from it could be shifted into Enron's broadband and power-trading empire. The spun-off company retained the upstream engineering team that had never been infected by Enron's culture — and over the next two decades became the technical pioneer of the American shale oil revolution. Today EOG Resources is worth roughly $70 billion in market cap.

Not shale. Not hydraulic fracturing. Not an oil-price rebound.
It was "getting thrown out of the house by the parent company's wrong strategy — and then the independent former engineering team using horizontal drilling plus staged fracturing to drag the technology of the North American shale oil revolution from the lab to the well pad."

EOG's predecessor was Enron Oil & Gas — the least conspicuous division inside Enron. Its CEO, Mark Papa — a geologist who happened to be spun off just in time before the Enron shipwreck and thus got to keep running the company — led EOG in the early 2000s as it became the first to apply horizontal wells traversing oil-bearing layers plus multi-stage fracturing to unlock crude in shale reservoirs that had previously been uneconomical to extract.

The Eagle Ford (Texas), the Bakken (North Dakota), and the Permian (West Texas) — EOG's early position-grabbing in all three of the most critical shale basins was among the most aggressive in the industry, and it stuck to "own rigs and self-developed completion fluid formulas" as its operational strategy. Papa refused to accept that "service companies can dictate our frac design." He said the chemical ratios in the fracturing fluid and the choice of proppant should remain in the hands of EOG's in-house engineers — because that was the core IP of shale extraction.

By 2024, EOG had at one point become the highest-valued independent upstream oil and gas producer by market cap on U.S. soil, surpassing ConocoPhillips and Occidental Petroleum. Papa still spoke on earnings calls in a geologist's tone — discussing the porosity and permeability of core samples, not Wall Street's discounted cash-flow rates.

Three things.
First thing: the old-economy division the parent company saw as "not flashy enough" — could be the next generation's dominant force.
Ken Lay thought oil extraction was "too traditional." He wanted internet bandwidth trading. He pushed the oil business away — and then that "old" division, after his death, grew up independently — into a colossus far larger than the wreckage of Enron.
In your company — is there a division that "looks old" — sitting neglected in a corner by management?
Second thing: fracturing fluid formula — it sounds like a farmer rendering pig fat — but it's actually the core IP of shale extraction.
EOG's differentiation wasn't "drill more wells." It was that its fracturing fluid, in every segment of shale fracture, had filling and conductivity efficiency slightly higher than the neighboring well's. That tiny margin, cumulatively, over the long run, represented billions of dollars in NPV.
Is your IP hidden inside a "doesn't look sophisticated" link in the production chain?
Third thing: don't outsource core operations — no matter how big the service company.
EOG insists on owning its rigs and its completion engineers — it does not give Halliburton or SLB teams full control. Because the true competitive factor is operational decision-making speed.
Is there a link you've outsourced — that is actually determining your competitiveness?

Mark Papa has been at EOG (and its predecessor Enron Oil & Gas) for over thirty-five years. He still goes personally every year to the Permian Basin and the Eagle Ford well pads to inspect cuttings samples. The line he repeats more than any other: "The drilling log is one thing — grabbing a handful of cuttings off the shale shaker and holding them in your palm is another."

← Previous Contents Next →
Chapter 167 Datadog

2010, New York City. Two French software engineers — Olivier Pomel and Alexis Lê-Quôc — sat in a cramped co-working space in Manhattan staring at server monitoring charts on their screens. They had just left Wireless Generation, a New York education-data company — where the two of them had managed operations and discovered that the tools for monitoring server and application performance — Nagios, Ganglia, Graphite, custom Perl scripts — were fragmented and abundant, and not a single tool could monitor cloud infrastructure and applications simultaneously. Pomel said to Lê-Quôc: "Let's not reinvent monitoring tools. Let's do one thing: ingest all the data and display it on a single screen." That tool would later shatter the fragmentation of cloud monitoring. Today Datadog is worth roughly $40 billion in market cap.

Not cloud. Not microservices. Not DevOps. Not AI.
It was "one dashboard that swallows it all" — server utilization, container health, slow database queries, application error logs, and network traffic — in the same interface.

Olivier Pomel was born in France in 1982 and earned a computer science degree from École Centrale Paris. Alexis Lê-Quôc studied mathematics and computer science at École Polytechnique — the two had already worked together at the same company back in France. After arriving in New York, they worked as software architects at the education-data company Wireless Generation — where they were tormented daily by fragmented monitoring tools. Servers used Nagios for alerts, Graphite to store time-series data, logs were scraped with custom Perl scripts — engineers spent huge chunks of their day jumping between different tools to trace the root cause of failures.

Datadog's first product had only one function: automatically discover EC2 instances and services on AWS — then aggregate all their performance metrics into a single customizable time-series dashboard. That "auto-discovery" made Datadog fundamentally different from every legacy data-center monitoring tool — it did not require manually configuring every monitored object. Because cloud instances spin up and shut down dynamically — manually updating a monitoring list was infeasible in an elastically scaling cloud environment.

Datadog subsequently expanded into three pillars: infrastructure monitoring, application performance monitoring (APM), and log management — all three sharing the same search query language on the same platform, with no need for an engineer to leave the current screen to dig through another tool. This "unified observability platform" saw demand explode during the remote-operations wave driven by the 2020 pandemic. Because when nobody was in the data center to check a blinking LED in person, data centers had to rely on dashboards like Datadog to "see" and "hear" the servers.

Three things.
First thing: you don't need a "new" feature — just consolidate the 7 existing tools into 1.
Datadog didn't invent monitoring. It took the functions of Grafana, Nagios, PagerDuty, Splunk, New Relic, and others — and collapsed them under a single UI. It's consolidation. Consolidation itself is a $40 billion product.
Are your customers suffering from "the switching cost between 7 tools"?
Second thing: cloud monitoring is a fundamentally different physics problem from traditional monitoring.
A traditional server sits in a data center — you configure the IP once and it stays unchanged for a year or two. A cloud instance — spins up today, auto-terminates 30 minutes later. Datadog's "auto service discovery" was built specifically to solve this dynamism. It didn't improve old monitoring — it redesigned for a new physical environment.
Is your product an "improvement on an old tool" or a "redesign for new infrastructure"?
Third thing: when monitoring, logs, and APM all share the same query language — the engineer never has to leave your dashboard.
When all the data lives inside the same search box — you don't need to "export data to Excel and then import it into Python." The workflow closes into a loop inside Datadog. The switching cost of leaving — becomes unbearably high.

Olivier Pomel now often appears in public wearing a hoodie printed with a Dash logo — Dash is the Datadog mascot, a cartoon dog. Playful — but Dash's real meaning in engineering culture is this: a dashboard that can keep up with the speed at which your code creates instances.

← Previous Contents Next →
Chapter 168 Phillips-66

2012, Houston. ConocoPhillips decided to spin off its downstream businesses — refining, chemicals, pipelines, and gas stations — into an independent company. The new company's name was chosen from a number in ConocoPhillips's history: Phillips 66. Originally just a blended-gasoline brand name launched in 1927 (Phillips + Route 66) — it was later repurposed as the new company's "old name." Today Phillips 66 is worth roughly $60 billion in market cap.

Not oil. Not shale. Not petrochemicals. Not brand marketing.
It was a number: 66. That asphalt road stretching all the way from Chicago to Los Angeles — the most famous highway in American history — Phillips, in 1927, put the name of Route 66 into its gasoline brand. From that moment, the "Phillips 66" sign carried, all at once, the imagery of an oil company, a gas station, American road culture, and the road movie.

In 1927, Phillips Petroleum — based in Bartlesville, Oklahoma — launched its first blended gasoline formula and named it "Phillips 66." The reason for the name: the company's test car had hit its top speed while driving on Route 66. The Phillips 66 highway-shield logo — white numerals on a red rounded-square background — has barely changed since.

In 2002 — ConocoPhillips (Conoco and Phillips merged in 2002) gradually consolidated its downstream chemicals, refining, and gas-station assets into an independent entity — and in 2012 fully spun it off as a separate public company. This newly independent company reclaimed the old name from 1927: Phillips 66. It is "an old soul in a new shell" — the new company is one of the world's largest suppliers of lubricant base oils, polyethylene, and refining catalysts — yet its brand still wears Route 66's old leather jacket.

Phillips 66 doesn't produce oil — it is a "converter" — transforming "crude that other people extracted" into gasoline, diesel, jet fuel, lubricants, and chemical feedstocks. Phillips 66's refineries are positioned downstream of America's most critical shale-producing regions: the Gulf Coast, the Midwest, the East Coast. Its joint-venture chemicals operation, Chevron Phillips Chemical (a 50/50 partnership with Chevron), is one of the world's largest producers of polyethylene and alpha-olefins.

Three things.
First thing: make the company name itself a landmark symbol.
Phillips 66 is not an ordinary brand name — 66 is a highway number. In America, the number 66 intrinsically carries the collective memory of "travel, freedom, crossing the continent." You don't need to explain to an American what "66" means — it's already in their brain.
Can your company name borrow a public symbol — instead of inventing a made-up word?
Second thing: spin off downstream — so upstream and downstream can each breathe inside their own oil-price quadrants.
ConocoPhillips split in 2012 — because the profit/risk model of upstream oil extraction and the profit/risk model of downstream refining are fundamentally different. After the split, shareholder expectations on both sides also better matched each business's own cadence.
Is your company squeezing "two different economic models" into a single financial statement?
Third thing: an oil company that doesn't drill — it only converts.
Phillips 66's competitive edge isn't "we extract oil more cheaply" — it's "we convert the oil others extracted more cleanly, more thoroughly, extracting more high-value product from every barrel." It's a "conversion efficiency" machine.

Phillips 66's headquarters sits in western Houston. In the lobby of the headquarters tower, there is a long "Route 66 roadmap" inlaid in brass across the floor — running all the way from Lake Michigan in Chicago to the Santa Monica Pier in Los Angeles. There are no tourists here — only oilmen walking over it on their way into meetings. Engineers in work boots passing through — habitually sidestep the Illinois section.

← Previous Contents Next →
Chapter 169 Ecolab

1923, St. Paul, Minnesota. A chemist named Merritt J. Osborn — who had previously researched dairy sterilization at the U.S. Department of Agriculture — stood in the banquet hall of a downtown hotel, pitching a "solid descaling agent for dishwashers" to the hotel manager. At the time, many hotel kitchens washed dishes with hot water — the temperature wasn't high enough, and it frequently led to food poisoning. Osborn's cleaning agent could break the chemical bonds of fats and proteins at lower temperatures. He named the company "Economics Laboratory" — shortened to Ecoclean — later called Ecolab. Today Ecolab is worth roughly $75 billion in market cap and is the world's largest provider of cleaning and water-treatment services for the restaurant, hospitality, healthcare, and industrial sectors.

Not chemistry. Not food safety. Not medical disinfection. Not water treatment.
It was one word that "every hotel kitchen has to carry out every single day": dishwasher cleaning solution. And then Ecolab, over the course of a century, extended the "cleaning" business laterally — to everything involving surfaces and water treatment, from disinfecting dairy-cow teats to recycling frac water from shale wells.

Osborn had originally been using chemistry to extend milk's shelf life — preventing foodborne illnesses caused by incomplete pasteurization. He began applying the same surface-disinfection logic to hotel dishes, kitchen knives, and stainless-steel countertops. During the Great Depression of the 1930s, hotels slashed costs — hot-water temperatures dropped from 180 degrees Fahrenheit to 140 — washing dishes at lower temperatures left massive bacterial residue. Osborn then developed a low-temperature degreasing agent that could dissolve grease at 120 degrees Fahrenheit — and sold it into the back kitchens of every chain restaurant of the era (such as Howard Johnson's).

Ecolab's truly revolutionary breakthrough came in the 1980s — it began selling "cleaning + monitoring service" packages. No longer selling buckets of chemicals — it sold "arranging a dedicated technician to inspect chemical concentrations at your restaurant on a regular schedule, install automatic dispensers, and conduct food-safety training." This was called "CIP (Clean-in-Place) service management." After you signed with Ecolab — your people never touched chemicals. An Ecolab technician came weekly to inspect and refill, and to record every food-safety regulatory compliance data point. This transformed the company from "a chemical company" into "a food-safety compliance and cleaning outsourcing service company."

Then Ecolab used the same logic to expand into healthcare (hospital surface disinfection and pre-operative skin treatment — acquiring the Microtek and Anios brands); oil and gas (post-frac water and fluid treatment — merging with Nalco to form the Nalco Water subsidiary); and industrial water treatment (anti-scaling and anti-microbial-growth chemicals for boilers and cooling towers). Ecolab now provides cleaning and microbial control for virtually every industrial and consumer scenario that involves "water, microbes, and surface contact."

Three things.
First thing: it's not selling cleaning agents — it's making it so others don't have to clean things themselves.
Ecolab turned "cleaning chemicals" into "cleaning service + weekly on-site inspections." A restaurant owner, after signing the contract — food-safety and hygiene compliance records generate themselves automatically — he never has to personally manage the cleaning crew.
Can your product upgrade from "selling tools" to "operating the tools for the customer"?
Second thing: from milk to industrial boilers — cleaning, at its essence, is two physical actions: "kill + wash."
Teats, countertops, surgical instruments, cooling towers, wellbores — they are all "surfaces that need to be cleaned" and "microbes that need to be killed or controlled." Ecolab exported this minimalist chemical capability onto nearly every kind of surface.
Can your core capability be applied to ten industries that look entirely unrelated to one another?
Third thing: cleaning is not a cost — it's "maintaining a hotel or restaurant's food-safety license."
A restaurant isn't afraid of cleaning solution going up by a dollar — it's afraid of a foodborne-illness outbreak shutting it down. Ecolab doesn't sell "cheap cleaning" — it sells "your establishment will never land on the front page of the newspaper because of a hygiene problem."
How much is your product worth — in terms of "loss avoidance" for your customer?

Merritt Osborn died in 1946. Today, in the lobby of Ecolab's global headquarters in St. Paul, beneath his name, are his original words: "As long as there is water on Earth — someone needs to know at what temperature the microbes in it will die."

← Previous Contents Next →
Chapter 170 Colgate-Palmolive

1806, Dutch Street, New York City. William Colgate — the 23-year-old son of English immigrants — opened a small workshop beneath a clockmaker's shop. He made only three things: soap, candles, and starch. The reason: all three required the same basic raw material — animal fat. He pushed the finished goods to the New York Harbor docks on a handcart and sold them to sailing-ship captains about to put to sea — because you couldn't bathe on board, you could only scrub with soap, and every ship had to check its inventory before departing. Today Colgate-Palmolive is the world's largest toothpaste and oral-care products company, worth roughly $75 billion in market cap.

Not dentistry. Not fluoride. Not oral-health promotion. Not the toothpaste tube.
It was "a ship captain's soap order form." New York in 1806 was the busiest port in America — every day, fifty or sixty Atlantic trading vessels took on fresh water and dry goods. Colgate walked in — asked the captain, "Do you have soap?" — and then memorized every ship's next expected return date. When that ship came back — he was at the dock, delivering the next week's soap.

William Colgate was born in England in 1783 and immigrated with his family to Baltimore at age six. His father ran a soap and candle workshop — William set up his own shop in New York at nineteen. In New York he started out boiling tallow and lye in a large iron kettle to make soap — every week he wheeled the finished product to the docks on a cart and sold it to ship captains, then saved the money to buy more animal fat and lye to expand the kettle.

By 1833, he was one of the largest soap and candle manufacturers in New York. His religious faith ran deep — before every bar was used, he would pray that "the person using this soap might be a little healthier because of it." Did this have anything to do with business? It did — because he insisted that every bar of soap be stamped with the "Colgate" trademark — because he guaranteed this bar was not counterfeit and had been produced to a fixed formula. This "brand faith" was later refined in 1896 by his son Samuel Colgate and his grandson — who launched the first tube of Colgate toothpaste — the first "toothpaste" ever packaged in a collapsible tin tube. (Before the 19th century, tooth powder was kept in jars — you dipped your finger or toothbrush into it, shared by the whole family — utterly unhygienic.)

Colgate toothpaste joined with Palmolive soap — the latter was launched in 1898 by the B.J. Johnson Company, the first soap made from palm oil and olive oil — and the two companies merged in 1928 to form Colgate-Palmolive. Then its oral-care line began a global colonization — from the United States to Latin America (especially Brazil and Mexico), to India, to Southeast Asia — preemptively "establishing the first-tube habit" in emerging markets where oral-health awareness was just beginning to rise.

Three things.
First thing: make a product — then show up before the customer's next need arises.
Colgate went to the docks not to take the day's orders — but to memorize the captain's next return date, make the soap ahead of time, and deliver it. This act, more than any later advertising, built "brand trust" — because the seller was there, waiting for you, before you even knew you needed him.
The next time your customer needs you — are you already standing in front of them before they open their mouth?
Second thing: from a jar of tooth powder the whole family dipped into — to "a personal tube of toothpaste."
Colgate moved tooth powder from a ceramic jar into a collapsible tin tube — creating the "personal toothpaste" category. Before, tooth powder was a family-shared item. Colgate said, "One tube per person." A physical packaging redesign — turned a hygiene habit into an individual consumer product.
Do you have a way to change consumer behavior through "packaging physics design"?
Third thing: use a short, memorable English word as a brand — and then spread the very act of "brushing teeth" into developing countries.
For a stretch of its marketing history in India, Colgate did something radically basic — it sent staff into villages to give children "brushing demonstrations." They weren't selling "a better toothpaste than the local brand" — they were selling "the habit of brushing itself." As long as you brush — use Colgate.

William Colgate died in 1857. His son Samuel donated a substantial portion of the family fortune to what is now Colgate University — in the valley of Hamilton, New York. The name has nothing to do with toothpaste — yet the overwhelming majority of the world sees "Colgate" on a subway or supermarket shelf and thinks toothpaste, and sees "Colgate" during the NCAA basketball tournament and thinks a college team.

← Previous Contents Next →
Chapter 171 Norfolk-Southern

1830, Charleston, South Carolina. The Charleston City Council, lobbied by local merchants, approved the issuance of the area's first railroad bonds — to fund the construction of a track running from Charleston to Hamburg, a route devoted entirely to cotton transport. The Charleston-Hamburg Railroad would become one of the earliest ancestors of the Southern Railway system. Then, in 1838, the Virginia General Assembly authorized the creation of the Norfolk and Petersburg Railroad Company — extending inland from the port of Norfolk, transferring port cargo onto rail and into the farmlands of Virginia and North Carolina. In 1982, the Southern Railway and the Norfolk and Western Railway merged — taking the name Norfolk Southern. Today NS is one of the two Class I freight railroads in the eastern United States, worth roughly $65 billion in market cap.

Not high-speed rail. Not passenger service. Not bullet trains. Not electrification.
It was two short 19th-century cotton-transport lines and a port-cargo transfer spur — pieced together 170 years later into an east-west network spanning from the Mississippi River to the Atlantic.

Norfolk Southern's original DNA is "port rail" — the connecting line between a harbor and inland farmland. The Port of Norfolk is Virginia's finest natural ice-free deepwater harbor — the Chesapeake Bay's tidal range is minimal, suitable for large-vessel berthing. In the 1860s the Norfolk and Petersburg Railroad moved Civil War military supplies and, after the war, the South's cotton, tobacco, and timber from the port inland to the rail junction at Petersburg — where cargo was handed off to other railroads for continued transport deeper into the interior.

The Southern Railway's Charleston-Hamburg line, meanwhile, connected South Carolina's agricultural production zones to the Port of Charleston during the cotton era. The legacy of these two railroads — both starting and ending at seaports — is deeply imprinted on NS's business today: it is one of the largest intermodal (container train) carriers in the eastern United States. Containers are unloaded from ships at the ports of Norfolk, Charleston, and Savannah — then loaded directly onto NS trains — pulled nonstop to Midwestern hubs.

NS is also a coal railroad — its Appalachian coalfield spurs haul thermal coal and metallurgical coal out of the mountains of West Virginia, western Virginia, and eastern Kentucky to power plants and export terminals. As America retires coal-fired generation, coal volumes have steadily declined — and NS is now repurposing the capacity on this "coal-line network" into a "logistics backbone" — moving e-commerce freight, auto parts, chemicals, and timber.

Three things.
First thing: rails — are universal infrastructure on which you can run completely different types of cargo.
In 1850, NS's tracks carried cotton bales and tobacco barrels. In 2025, they carry shipping containers and auto parts. The value of the rails themselves — is reused across different industrial cycles, again and again.
Your "rails" — that underlying foundation — can it support entirely different "cargo"?
Second thing: port railroads — turning the last few miles of the sea-land interface into a closed-loop toll.
The container ship is the world's greatest cargo-aggregation model. A ship unloads ten thousand containers at the Port of Norfolk — and NS loads those boxes straight onto trains — no intermediate transload warehouse — direct pull to inland intermodal terminals. The efficiency of this "port to train" handoff is the core moat of the port railroad.
Is your business positioned at "the interface between two transport/transmission and transfer nodes"?
Third thing: use mergers to consolidate a fragmented network — then unify the operating system.
NS was assembled segment by segment from dozens of originally independent small railroads. They once had different rail weights, different signaling systems, different interchange protocols at their boundaries. What NS did was unify them all into a single operating system — a single dispatch center.

Norfolk Southern's old Norfolk terminal building — built in 1903 beside an inland-waterway berth — is now the office of a trading company. The old narrow-gauge rails are still buried beneath the floor — and whenever a heavy-haul train passes by on the main line a few miles away, those long-dead rails still emit a faint low-frequency hum.

← Previous Contents Next →
Chapter 172 Digital-Realty

2004, San Francisco. A real-estate investor named Michael Foust, who had successfully managed the data-center assets of GI Partners (a middle-market private-equity fund), decided to spin those data-center assets out and take them public separately. He named the new company "Digital Realty Trust" — dedicated to owning and operating data-center buildings. In 2004, the overwhelming majority of companies were still building their own server rooms — "data-center landlord" as a category barely existed. Foust said: "The day will come when every company moves its servers out of the basement — and into our buildings." Today Digital Realty is one of the largest data-center REITs in the world, worth roughly $60 billion in market cap.

Not cloud computing. Not AI. Not internet traffic.
It was "servers don't live in your office basement — they live in purpose-built buildings, and you pay rent for that."

During his time at GI Partners, Foust acquired a batch of aging telecom switching centers and corporate data centers at low prices — assets that were being dumped by their original owners for fire-sale prices after the dot-com bubble burst — because the internet companies had gone bankrupt. But what he saw wasn't "old switch rooms nobody wants anymore." He saw that inside these buildings, the physical infrastructure for carrying enormous amounts of power and cooling was already in place: total power capacity (MW), cooling towers (chillers), diesel generator sets, and independent feeder lines — in other words, the "data-center shell." To put new servers into these shells, you didn't need to construct a new building — just plug in.

Digital Realty's business model is that of a REIT (Real Estate Investment Trust) — buy land or acquire long-term leaseholds, construct data-center buildings, and lease them out under long-term contracts to tech companies, cloud providers, banks, and government agencies. Its rent is calculated by power capacity (per kW or per MW) — not by square footage. This is different from office-building leasing — DRT collects money for every watt of electricity.

In the 2010s, cloud computing exploded — AWS, Azure, and Google Cloud needed to deploy hundreds of edge nodes (data centers close to users) around the globe to reduce latency. Digital Realty's global footprint made it the natural "data-center landlord" for cloud providers — in many cities it already owned the primary internet exchange buildings for the region. With the explosive growth of AI model training — GPU clusters require colossal power consumption and liquid cooling — Digital Realty is now upgrading some of its older data centers to support cooling capacity of 50 kW or more per rack.

Three things.
First thing: after the dot-com crash — salvage the "shells" of old switch rooms.
DRT's earliest assets were empty machine rooms left behind by bankrupt telecom companies — but the power was still there. The generators were still there. The cooling towers were still there. It acquired the physical remains of shipwrecks at rock-bottom prices — and leased them out to the next wave.
In your industry — is there "physical infrastructure" abandoned by the last bubble — that you can acquire cheap and flip?
Second thing: collect rent — per watt, not per square meter.
Data-center tenants care about how many kilowatts of servers they can fit — not how many square meters of desks they can place. Digital Realty's rent model is priced by capacity — aligned with the logic of the industry.
Are your pricing units aligned with your customer's value units?
Third thing: the data center of the future is no longer "a big warehouse with cold air" — it's "liquid-cooling lines running through every rack row."
AI chip power density has already surpassed 50 kW per rack — traditional air cooling cannot dissipate that heat; liquid cooling or immersion cooling is mandatory. Digital Realty is upgrading the cooling systems in some of its older buildings to accommodate AI's demands. Its buildings aren't "buy and forget" investments — they undergo rapid technical retrofitting driven by tenant needs.

Michael Foust stepped down from his management role after 2018. But in the first investment journal he kept when founding DRT — on one page he drew a box, annotated beside it: "This box consumes power and exhausts heat every hour of every day. Someday, every company will need to put its box — inside my box."

← Previous Contents Next →
Chapter 173 Royal-Caribbean

1968, Oslo, Norway. Three Norwegian shipowners — Arne Wilhelmsen, Sigvald Iversen, and Anders Wilhelmsen — sat in their ship-brokerage office, gazing at the dismal passenger numbers on transatlantic liner routes, and said: "Nobody wants to cross the Atlantic by ship anymore. Flying across takes just seven hours. We have to build something at sea that flying can't reach." They converted three old ferries into "floating resort hotels" with no destination port — the ship wasn't a means of transport from A to B; the ship itself was the destination. Today Royal Caribbean is worth roughly $70 billion in market cap and operates the largest passenger ships ever built in human history, including Icon of the Seas.

Not tourism. Not hotels. Not an aviation substitute. Not the Caribbean Sea.
It was the counterintuitive idea of "turning the ship from a means of transport into the destination." When everyone was abandoning ships for planes, these three men cut more balconies into the hull, put in swimming pools and buffet restaurants — and sailed for Miami.

Arne Wilhelmsen was born in Norway in 1921 — a country renowned in global shipping for "good ships and good shipowners." His early companies were primarily in bulk shipping, carrying general cargo and oil — thin margins, deeply cyclical. By the late 1960s, Wilhelmsen and his two partners noticed something: in the Caribbean, Florida — especially Miami — a new kind of tourist flow was emerging. Tourists from the cold northern states were willing to fly to warm tropical port cities and board a cruise, provided the ship had enough "fun things" on it — which particular islands it visited didn't much matter. They subsequently founded Royal Caribbean Cruise Line.

In 1970, Royal Caribbean's first new-build ship, Song of Norway, set sail on her maiden voyage — designed from the start with the open deck doubled in size, and featuring a circular bar wrapped around the funnel amidships called the Viking Crown Lounge — the most distinctive architectural landmark at sea. After that, the company added roughly one ship per year, gradually expanding.

In 1988, Royal Caribbean and its chief rival, Miami-based Carnival Cruise Line, entered a "megaship arms race" — from 1988's Sovereign of the Seas (73,000 gross tons — hailed at the time as "Sovereign of the Seas") to 1999's Explorer of the Seas (137,000 tons) to 2009's Oasis of the Seas (225,000 tons — the first to introduce "Central Park" and a promenade-style shopping street). Today the newest Icon of the Seas (maiden voyage 2024) exceeds 250,000 tons and includes a water park.

Three things.
First thing: aviation devoured all transoceanic passenger shipping — and the ship survived because it stopped racing planes on speed.
Royal Caribbean watched the Boeing 707 steal every transatlantic ship passenger — and then said: "We aren't transporting people anymore. We're building floating playgrounds with no destination."
When your industry's core function gets replaced by another technology — are you willing to give up that function and invent a new one?
Second thing: "building the ship" itself — is the competitive barrier.
The shipbuilding contract for Icon of the Seas exceeded $2 billion. Building a ship like this takes at least three to four years. No new entrant can possibly have the shipyard capacity or the capital in the short term. Royal Caribbean's moat isn't "patents" — it's that every new ship is physically impossible for a competitor to replicate quickly.
Can your industry's physical production cycle — become your competitive moat?
Third thing: the cruise revenue model — expanding from the ticket to "average onboard spend."
Royal Caribbean now earns more from onboard casinos, alcohol beverage packages, water park admission, specialty restaurant reservations, internet data packages, and shore excursions — than from the ticket itself. Once a passenger boards — every optional consumer product is in the shipowner's hands.

Arne Wilhelmsen died in 2019 — on the aft bar deck of Sobrena of the Seas, there still sits a cut section of the original teak decking from the very first Royal Caribbean ship. That ship was later sold to a small Greek passenger line. Only those few planks of wood made it back to Miami.

← Previous Contents Next →
Chapter 174 Robinhood

2013, Palo Alto, California. Two Stanford graduates in math and physics — Vladimir Tenev and Baiju Bhatt — sat in a coffee shop, surrounded by dozens of printed-out pages of SEC rule filings. They had discovered something: the back-end execution cost of a U.S. stock trade had already dropped to zero — yet every traditional brokerage was still charging $0.01 per share in commissions. Tenev said: "You can get stock quotes for free on your phone — why does it cost seven dollars to buy a hundred shares of Apple?" Then Robinhood was born — the world's first zero-commission stock trading app. Today Robinhood is worth roughly $50 billion in market cap.

Not financial innovation. Not high-frequency trading. Not cryptocurrency. Not options trading.
It was "zero commission." Three paradigm-shifting words — that crushed the entire U.S. retail brokerage fee model to zero.

Tenev was born in Bulgaria in 1987 and moved with his parents to the United States at age five. He met Bhatt while studying math at Stanford. After both worked on Wall Street, they discovered that the back-end marginal cost for a brokerage to execute a stock trade was approaching zero — because automated clearing and electronic exchanges had fully matured by the 2010s. Traditional brokerages charging $7 to $10 in commissions weren't covering costs — they were covering their physical branch offices and sales staff.

In March 2015, Robinhood officially launched, offering zero-commission trading. In its first week, the waitlist surpassed 500,000 email addresses. The traditional brokerages responded with mockery — "zero commission cannot be a sustainable business model." Robinhood's answer was the payment for order flow (PFOF) model — it sold its clients' orders to high-frequency trading firms (such as Citadel Securities), which paid Robinhood rebates for the additional order flow. This quietly ignited an ongoing controversy over whether clients were truly getting best execution.

In January 2021 — the GameStop retail frenzy — Robinhood temporarily restricted buying of GME and certain meme stocks after its clearinghouse's intraday margin requirements spiked, triggering bipartisan outrage across the global political spectrum and a wave of angry hearings. Tenev was grilled repeatedly before Congress. But Robinhood's user count continued to climb afterward. Robinhood then went public in the summer of 2021 — under the ticker HOOD.

After 2023, Robinhood expanded into Europe and the UK, adding cryptocurrency trading, retirement accounts, credit cards, and predictive-contract trading — gradually transforming from a "retail stock-trading app" into a "universal personal finance terminal."

Three things.
First thing: when the back-end cost approaches zero — the front-end price should be zero.
Robinhood's emergence is a warning to every fintech founder: if your competitor's back-end marginal cost is already nearing zero — but they're still charging according to the "old pricing structure" — that's a new entrant waiting to shatter the profit pool with "zero."
In your industry, has the "back-end cost" already approached zero? If so — can your price be zero?
Second thing: payment for order flow is a business model — but not necessarily an ethical one.
Robinhood's zero commissions are funded by PFOF rebates — but PFOF means client orders can be previewed by high-frequency traders at microsecond speed. After 2020 the SEC launched a deep market-structure review.
Is your revenue source — the part the customer doesn't know about?
Third thing: turning investing into a red button and a green button on a phone — the power of democratizing financial products.
Robinhood's minimalist interface — buy and sell, just two buttons — brought a vast number of people who had never bought a stock into the stock market for the first time. Its design philosophy is "delete every technical term and every table" — retaining customers by "removing every barrier to entry."

After the GameStop hearings, Vladimir Tenev has almost never spoken publicly about politics. He still rents a place in Palo Alto — his largest personal asset is Robinhood stock. At an internal all-hands meeting, he once said: "Our customers are not Wall Street's — they're the people who, until now, have been locked out of investing."

← Previous Contents Next →
Chapter 175 Ross-Stores

1950, San Bruno, California. A buyer who had worked for years in San Francisco department stores — Morris "Morrie" Ross — noticed a strange fact: every large department store and branded boutique had an "inventory black hole" — last season's unsold clothes, odd-size shoes from broken sets, and cosmetics returned with damaged packaging. These goods occupied shelf space that should have been displaying the newest arrivals — but since they wouldn't sell, they got shoved into a corner under a "clearance" sign. Ross opened the first Ross Dress for Less on 29th Street in San Francisco — sourcing directly from the "excess inventory" channels of brands and department stores and selling at 20% to 60% below original retail. Today Ross Stores is worth roughly $50 billion in market cap.

Not apparel. Not fast fashion. Not e-commerce. Not luxury.
It was "disposing of other people's inventory black holes" — turning department-store dead stock and broken-size runs into its own entire shelf assortment.

Morris Ross was born in the 1920s to a Jewish family in San Francisco. He started from the bottom as an assistant buyer at a department store — learning how to read inventory turnover ratios and end-of-season markdown data. He observed: every department store had 10% to 15% of its seasonal inventory that ultimately got liquidated to jobbers or discount companies at prices far below cost — because they had no space to store the next season's merchandise.

In 1950, Ross opened the first Ross Dress for Less — not in a prime location, but in a suburban area near residential neighborhoods — the rent was cheap. The shelves had no fixed brands — you didn't look at the brand, you looked at the buying opportunity. Today's shipment might have Polo shirts, tomorrow's might have Calvin Klein jeans, a week later a batch of kitchen towels offloaded from JCPenney. Customers didn't walk in expecting "the same thing every time" — they expected "a surprise every time." This is "treasure-hunt shopping" — TJX Companies (parent of TJ Maxx and Marshalls) uses the exact same model. Ross and TJX are the off-price duopoly in America.

In 1982, Ross Stores was taken over at the helm by a man named Stuart Moldaw — Morrie Ross's son-in-law. Moldaw expanded the company from California across the entire country — over thirty years, growing the store count from 35 to more than 2,000, starting from scratch. He never altered the sourcing model: "Ross's buyers have to be the people in the entire industry who understand 'dead stock' better than anyone. Others buy the 'next season.' We buy 'last season that nobody wanted.'" Ross does not do e-commerce — because its inventory only works when "the customer stands in front of the shelf — and this item is the only one, right here, right now." It can't be displayed online — because each batch is too small, and the cost of photographing and creating product detail pages would eat all the profit.

Three things.
First thing: don't sell what's trendy now — sell other people's dead stock.
Department stores and branded boutiques have to maintain a fashion image — they must clear inventory every season. Ross feeds on that "clearance" — at discounts of more than half off the original retail. But the cost price is even lower than the original cost — so Ross's low selling price still carries positive gross margin.
In your industry — is there "last season's inventory" that nobody can handle themselves — where you could become the "digestive black hole"?
Second thing: not doing e-commerce isn't about technology — it's because the physical item is only in the store.
Ross might have only a few units of any given item — scattered across different stores. You can't use warehouse logic to do Ross online — because there's no way to produce the polished photos an online listing requires for every SKU.
Is your channel strategy — aligned with the "uniqueness" of your inventory?
Third thing: low-rent suburbs — open near residential areas.
Ross doesn't pay premium shopping-mall rent — it leases side-wing spaces in suburban shopping plazas. Its customers are neighboring residents who "drive five minutes to browse — and may or may not buy." The cost of that foot traffic is extremely low.

Morris Ross died in 1995. Stuart Moldaw stepped down as CEO in 2014. When asked whether off-price could be replaced by e-commerce, Moldaw gave a line that has been quoted again and again: "E-commerce can't sell the last pair of odd-size size-6 maroon-purple sneakers on your shelf to a woman who happens to wear a size 6 — and have her walk out of the store wearing them right then."

← Previous Contents Next →
Chapter 176 Comfort-Systems-USA

1997, Houston. A mid-sized HVAC (heating, ventilation, and air conditioning) contractor — formed by engineers and project managers who had split off from larger mechanical contracting firms — went public on the Nasdaq. It acquired dozens of localized HVAC, electrical, and plumbing contracting shops and consolidated them under a single operating platform. Today Comfort Systems USA is one of the largest commercial HVAC and building mechanical-electrical installation service companies in the United States, worth roughly $25 billion in market cap.

Not HVAC technology. Not construction cycles. Not data centers. Not the physics of refrigeration.
It was buying hundreds of family-run HVAC installation and repair shops across America — the kind that could only take on small local projects — giving them the same procurement system, the same safety training, the same ERP and project-pricing model — and turning a pack of "little workshops" into a group capable of bidding on massive data-center and semiconductor-fab cooling systems across dozens of states, all at once.

HVAC is one of the most fragmented industries in American building technical services — more than 80% of HVAC companies are family-run, with just a single office. From its IPO in 1997, Comfort Systems began systematically acquiring leading local HVAC contractors — keeping the original names and original managers (because they had decades-old relationships with local general contractors and suppliers) — but unifying the back end: materials procurement, insurance, safety protocols, and project-tracking software.

When data centers, chip fabs, and electric-vehicle battery factories began being built at a massive scale in the United States in the late 2010s — the cooling capacity and specialized cleanroom HVAC precision that these industrial facilities demanded had already exceeded the design and construction capability of local small shops. And Comfort Systems' national footprint and unified engineering capability made it one of a very small number of qualified contractors capable of taking on these "super cooling system" projects.

Three things.
First thing: acquire fragmented local family HVAC shops — keep their signage and their local face.
HVAC is a local-relationship business — the general contractor trusts the old technician he's known for twenty years. Comfort Systems didn't change those family companies' names — it only unified the back-end ERP, insurance, safety, and materials procurement.
When you acquire — do you kill the acquired brand's local face, or preserve it?
Second thing: cooling a data center isn't about installing air conditioners outside the building — it's about designing the entire thermal cycle inside.
Comfort Systems' leap in competitiveness came from going from "fixing air conditioning for office buildings" to doing the full thermal design — from cold source to cooling tower to under-floor air distribution — for hyperscale data centers. The complexity multiplied dozens of times — the contract value multiplied a hundredfold or more.
Is there something in your "old customer needs" that is becoming extremely complex — where you could upgrade from "repair" to "design"?
Third thing: acquisition and consolidation — slash the "materials cost."
Hundreds of local HVAC companies each buying copper tubing, refrigerant, and fan coil units separately — unit prices were extremely high. Once consolidated under a single group — unified bulk purchasing drove materials costs down 5% to 10% — a colossal advantage in low-margin construction contracting.

Comfort Systems' headquarters is in Houston — no flashy elements whatsoever — behind the building is a row of welding training rooms and AutoCAD classrooms for training new hires and managers. The sign at the front desk reads: "We don't make cold — we remove heat."

← Previous Contents Next →
Chapter 177 DoorDash

2013, a classroom at Stanford University. Four Chinese and Chinese-American students — Tony Xu, Andy Fang, Stanley Tang, and Evan Moore — were doing research for a small class project on "digitizing local small businesses in Palo Alto." They walked into a dessert shop — the owner flipped open a stack of papers on the front counter to show them. That stack of papers was "delivery orders" — not electronic orders — scribbled phone messages. Every call, a customer read the menu and the owner wrote it down by hand. Then the owner sent his teenage son to drive and deliver. DoorDash's starting point was that stack of handwritten phone messages. Today DoorDash is worth roughly $70 billion in market cap.

Not food delivery. Not logistics. Not the on-demand economy. Not local commerce.
It was "that stack of notepad paper on the restaurant's front counter." Manual, error-prone — one wrong address and the order was gone for good. The first code Tony Xu and his co-founders wrote — called PaloAltoDelivery.com — was simply an extremely bare-bones menu page for that dessert shop and seven or eight other Palo Alto restaurants. No app. You could only order from a webpage. They drove the deliveries themselves. At most, a few orders a day. But that stack of phone-message paper — disappeared.

Tony Xu (Xu Xun) was born in Nanjing, China in 1990 and immigrated with his parents to Illinois at age five. He met the three other co-founders while studying industrial engineering at Stanford. The Stanford project was originally just a class assignment — but after the project wrapped up, they discovered restaurant owners kept calling them: "That website of yours — can you please not turn it off?"

In the summer of 2013, none of the four went to work at any Silicon Valley company — they entered Y Combinator and went full-time on DoorDash. At YC's Demo Day, what they showed investors wasn't a fancy business model — it was the comparison: that dessert shop's order volume before DoorDash (phone + handwritten) and after (webpage + auto-confirmation) — it had doubled.

DoorDash would later surpass Grubhub, Uber Eats, and Postmates to become the number-one food delivery platform in the United States by market share — primarily through its "suburb strategy." Instead of waging an expensive war with Uber Eats in dense urban cores, it aggressively entered the suburbs that most delivery platforms deemed "not worth delivering to" — where there was almost nothing besides Chinese takeout. DoorDash built "city-wide delivery" — any distance, any storefront, as long as it had a kitchen, could take orders.

The 2020 pandemic sent DoorDash's revenue — soaring overnight. It went from a company still fighting for survival to the basic survival pipeline for the restaurant industry. At the end of 2020, DoorDash went public — its market cap shot straight past $70 billion.

Three things.
First thing: don't think about the App — first get in your own car and deliver dozens of orders.
In the early months, Xu and Fang personally delivered dozens of orders a day — to know what drivers faced: hunting for parking, waiting for food, wrong customer addresses, not enough change. They learned every bit of friction from delivering — the kind that would later inform all of the product design.
The product you build — have you used it yourself, over and over, until your feet hurt and your hands went numb?
Second thing: the suburbs — not a remote wasteland that drags down margins — but a gold mine nobody wanted to touch.
Early on, DoorDash avoided the restaurant-dense zones of downtown San Francisco already saturated by competitors — it went to the suburbs south of San Jose, to Pleasanton, to Danville — delivering Chinese food, pizza, fried chicken. The customer options there were extremely limited, average order values were higher, and user churn was lower.
Is your market — in the "suburbs" everyone else is ignoring?
Third thing: build your own logistics network — not just an "order-matching" platform.
From the start, DoorDash built its own driver network, not just an "ordering page" — because this would later let it offer more controllable delivery times than Grubhub. This choice to "build your own capacity" — is what let it maintain control even when facing old rivals like Uber Eats with global fleet experience.

After 2023, Tony Xu became less involved in daily operations — but he still orders delivery via DoorDash at random times and personally delivers a few orders — under a fake name — testing the driver experience. He says: "The easiest thing for a boss is to con himself into thinking 'we don't need to deliver anymore ourselves.' And then you make mistake after mistake — and you can't even feel them."

← Previous Contents Next →
Chapter 178 General-Motors

1908, Flint, Michigan. William C. Durant — a self-taught businessman who had become one of the wealthiest men in Michigan by manufacturing horse-drawn carriages at the close of the 19th century — bundled together Buick, Oldsmobile, Cadillac, and several other independent automakers and formed a company called General Motors. His core logic wasn't building cars — it was using a parent company's stock to acquire more car brands. Today General Motors is worth roughly $60 billion in market cap.

Not automobiles. Not the assembly line. Not the internal combustion engine. Not electric vehicles.
It was Durant alone — a man who founded GM twice in his lifetime and was thrown out of the board of the company he created, both times. His method was "using a stock premium to acquire brands at different price bands — then stitching together total coverage of all customer tiers, from cheap to luxury." This logic was later adopted by Ford and Chrysler, too — but Durant was the first to pull it off.

Durant was born in Boston in 1861 and raised in Michigan. He began working at his grandfather's lumber mill as a teenager — then moved on to selling cigars, fire insurance, and real estate. In 1885, he and a friend bought a small stake in a Flint wagon manufacturer for $50 — and quickly turned the company (the Flint Road Cart Company) into one of the largest carriage makers in America. But Durant himself — initially refused to learn to drive. Even while running Buick, he still rode to the factory every day in a horse-drawn carriage.

In 1908 he merged Buick (mid-range), Oldsmobile (upper mid-range), and Cadillac (top tier) — forming General Motors. Afterward he kept adding brands — Chevrolet (founded by race-car driver Louis Chevrolet and later bought by Durant) was the critical brand for reaching ordinary-income buyers.

But Durant was terrible at finance and cash-flow management. Twice, due to overexpansion and collapsing stock prices, he was forced out of the GM CEO role by bankers and investors — including agents of the du Pont family and J.P. Morgan — and was ultimately compelled to leave GM entirely in the 1920s. His later years were so bleak that he made a living operating a bowling alley. When he died in 1947, he was nearly penniless.

After Durant's departure, GM under Sloan's management invented "planned product obsolescence" (minor model tweaks every year to stimulate trade-ins) and "multi-brand ladder pricing" (Chevrolet ← Pontiac ← Oldsmobile ← Buick ← Cadillac), which made GM the dominant force in the global auto industry from the 1930s through the 1970s. But in the 21st century, GM endured its 2009 bankruptcy and government bailout, the 2014 ignition-switch defect scandal, and the wrenching pain of the electrification transition.

Three things.
First thing: a man founded a company — and then was thrown out the door of the company he built with his own hands. Twice.
Durant was forced out two times — because he was a genius acquirer and shareholder-persuader with absolutely no ability to run a large-scale operation or control spending. The founder and the operator do not have to be the same person — but the board has to know how to switch between them.
Can you accept, right now, going from "founder" to "former founder"?
Second thing: a minor redesign every year — making car owners psychologically auto-replace their cars.
Sloan's "annual model change" strategy is the ultimate definition of consumerism — the car isn't undrivable, but next year's model has different turn signals — so your car "looks old."
Can your product — through "minor visual upgrades" — trigger a self-obsolescence impulse in your consumer?
Third thing: bankruptcy is not the death of a company — it's a handover between creditors and shareholders.
GM filed for bankruptcy protection in 2009 — the U.S. government bailed it out with taxpayer money — and restructured it as New GM. Old GM's stock became waste paper. New GM relisted in 2010 — new shareholders profited, old shareholders went to zero.

When William Durant died in 1947, his obituary read: "He closed his eyes to the sound of bowling pins falling. One street over — hundreds upon hundreds of Buicks, Chevrolets, and Cadillacs were rolling down the Michigan streets. Not one of them bore his name."

← Previous Contents Next →
Chapter 179 Warner-Bros-Discovery

1923, Hollywood. Four Polish-Jewish immigrant brothers — Harry, Albert, Sam, and Jack Warner — rented a vacant cold-storage meat locker on Gower Street and converted it into a film-screening room and office. They had only one old camera and a horse that could perform tricks in front of the lens — Warner Bros.' first film was a dog-and-horse show. In 1927 — at the Warner Theatre on Broadway in New York, Sam Warner played back the first synchronized spoken line: "Wait a minute — you ain't heard nothing yet!" — the ad-libbed words of Al Jolson in The Jazz Singer, the first feature-length talking picture in history. Today Warner Bros. Discovery is worth roughly $30 billion in market cap.

Not movies. Not television. Not streaming.
It was "Wait — you haven't heard anything yet!" — putting the human voice track into film. That ad-libbed line would later become Warner Bros.' DNA: always being the first to put a new technology that the industry considered "not worth investing in" into the content business — from talkies to color film to television to streaming.

The four Warner brothers were born between 1890 and 1900 in Ontario, Canada (their father had moved from Poland to Canada and then to the United States) and grew up in Youngstown, Ohio. Jack Warner, the youngest brother, was the soul of the company at the helm. In the 1920s they rented a small building in Hollywood — shot by day, developed the film themselves by night.

In 1926, Warner Bros. released Don Juan, the world's first film with a synchronized musical score accompanying the silent image — reviews at the time said, "When the music syncs with the picture — you can't help yourself." The other major studios (Paramount, MGM) believed talking pictures would destroy the "universal language" of silent cinema — and refused to follow. Then in 1927, Warner Bros. proved with The Jazz Singer that sound does not destroy film — sound triples the box office overnight.

The decades that followed saw Warner Bros. known for a "hard style" — crime films, World War I pictures, urban noir, and Western epics. Casablanca (1942) was the accidental perfect intersection of wartime propaganda and romantic drama. Then, in the 1950s, Warner Bros. aggressively entered television production — supplying series programming to ABC and other new TV networks.

In 2022 — the AT&T-era WarnerMedia merged with Discovery to form today's Warner Bros. Discovery. The merger saddled the company with a massive debt load — but also brought HBO, DC, Harry Potter, and Discovery's unscripted documentaries together under a single streaming brand, MAX. This "century-old content power station" is now locked in a battle with Netflix and Disney for the third spot in the global streaming market.

Three things.
First thing: bet in advance on a single technology — and then change the fundamental rules of the industry.
In 1926-27, Warner Bros. wagered its entire capital on talking pictures — and once it succeeded, every movie theater was forced to retrofit sound equipment, and silent films were permanently eliminated.
In your industry — is there a technology decision where "once you bet first — everyone else has no choice but to follow"?
Second thing: don't build your own chain of movie theaters — pour the capital into "making more and bigger pictures."
Unlike Paramount, Warner Bros. didn't sink money into cinema real estate — it concentrated its capital on production. When the Supreme Court later forced studios to divest their theaters — Paramount was dismantled — Warner was barely affected.
Are your assets allocated toward "distribution" or "creation"?
Third thing: a century of content IP — Harry Potter, the DC Justice League, HBO's "televised poetry" — is an asset portfolio.
Even after mergers, acquisitions, and more mergers — those names (Batman, Superman, Voldemort, the dragons of Game of Thrones) all carry a vast future of derivatives and fan bases.

Jack Warner died in Los Angeles in 1978. He was the last of the four brothers to pass. He was once asked, "Why did Warner Bros. make so many gangster films?" He shrugged and answered: "Because we brothers grew up knowing those people on the street. We weren't making movies — in our lives, we knew every person who could turn 'watching a movie' into a real shooting."

← Previous Contents Next →
Chapter 180 Energy-Transfer

1996, Dallas, Texas. Kelcy Warren — an engineer who had once run a small pipeline-welding and natural-gas processing plant in Texas — started with just $20 million, buying a narrow, old Texas intrastate natural gas liquids pipeline at a bankruptcy auction, a line that was barely transporting anything anymore. Warren said to his partner: "As long as Texas is still producing gas and oil every day — this dead pipe isn't dead. Connect it to the downstream plants." The company he founded, Energy Transfer, is worth roughly $70 billion in market cap today.

Not energy. Not pipelines. Not a producer. Not LNG.
It was connecting existing oil fields to existing processing plants to existing export terminals — stitching the three together by repurposing old, neglected pipelines into a single seamless network.

Kelcy Warren was born in 1964 in Texas to a pipeline-worker family. Right out of high school he was in the ditches — welding, laying, pressure-testing, pigging. What he saw in 1996 was this: the American natural gas and liquids pipeline industry was riddled with "orphan pipelines" — they had once belonged to a major company, part of a connection between a now-shuttered refinery and an oil field — but the pipe itself could still hold pressure. It just needed to be reconnected at both ends.

Warren's distinctiveness lies in this — unlike Kinder Morgan or Enterprise Products, he didn't chase the most economically optimal layout or the shortest route. His style was "buy up every leftover connection that looks like junk" — taking a heap of abandoned and unrelated pipes — through patching and mending — welding them into a patchwork quilt capable of serving all kinds of flow: Marcellus shale gas heading north, Permian heading west, the Mont Belvieu NGL and liquids hub.

After 2005, Energy Transfer, through large-scale M&A and new construction — most notably the acquisition of Sunoco's pipeline division and the construction of the Dakota Access Pipeline (DAPL) — became one of the largest natural gas and natural gas liquids pipeline networks in the United States. DAPL, because it crossed Native American protected lands, became a global flashpoint of environmental protest — Energy Transfer's legal and physical battle with protest groups and the federal government dragged on for years and ultimately ended with oil flowing through the pipe.

Three things.
First thing: you don't need the best assets — you need enough "connections."
Warren's very first pipeline was one someone else had abandoned. But that dead pipe happened to sit between a growing natural gas processing plant and a string of new shale wells. The same pipe — on the old map, it was a derelict line — on the new shale map, it was a shortcut.
Is your old asset — worthless on the "old map" but a critical thoroughfare on the "new map"?
Second thing: patch and mend — don't build from scratch.
Energy Transfer's acquisition habit is — buy existing low-flow or abandoned pipelines, then reverse the flow direction, add compressor stations, swap in larger-diameter pipe sections. Digging a brand-new pipeline from scratch is enormously expensive and requires federal and multi-state environmental approvals — patching a section of old pipe is far faster.
Does your speed at "patching" old things exceed your competitors' speed at "building new"?
Third thing: a pipeline — is a transport route — and it's also a "physical long-term contract."
Once oil and gas start flowing through a pipeline every day — the upstream producer won't easily switch pipelines — because they've already sunk drilling and fracking investments into that gathering system. The pipeline is a "physical lock," not a "contractual lock."

Kelcy Warren remains the Chairman and CEO of Energy Transfer. His office displays a section of old steel pipe — a cut from that first narrow, abandoned line from 1996. On the outside wall of the pipe, the previous company had spray-painted in block letters: "1993 — ABANDONED." Warren says: "Our story is — 'abandoned' doesn't mean 'empty.'"

← Previous Contents Next →
Chapter 181 Rocket-Lab

2006, Auckland, New Zealand. A New Zealander who never went to university — Peter Beck — began hand-mixing rocket fuel and building small solid rockets in a corrugated-iron shed in his backyard. No NASA. No government contracts. No space agency. What he had was an old lathe his mother no longer wanted and a second-hand milling machine. With these he built his first rocket engine — less than 1 kilonewton of thrust. The company that followed, Rocket Lab, is now the second-most prolific rocket company in the world for launching small satellites into orbit — second only to SpaceX. Today Rocket Lab is worth roughly 20 billion dollars.

Not aerospace. Not defense. Not smallsats. Not 3D printing.
It was Peter Beck. One man in a backyard, feeding liquid oxygen and kerosene through a stainless-steel injector he built himself into a combustion chamber — and ultimately turning the concept of a "dedicated small orbital rocket" from "never going to happen" into "launching every two weeks."

Beck was born in 1974 in rural Invercargill, New Zealand. His father was a museum exhibition designer, his mother worked at a school. He was fascinated by "up" from an early age — in primary school he packed gunpowder into iron pipes and tried to launch them over the school sports field. After high school he went straight to work as a toolmaking apprentice at a nearby mineral-washing plant — where he learned to cut and weld stainless steel and titanium alloys. By his thirties he was a materials and engineering specialist at Callaghan Innovation, a New Zealand government research institute — and he still had no university degree.

Between 2006 and 2013 — working through his New Zealand company and later Rocket Lab, registered in the United States — he designed, built, and tested over a dozen small rocket engines entirely on his own. His first New Zealand rocket company was staffed almost entirely by self-taught materials engineers and former auto mechanics. In 2013 he decided to move headquarters to Huntington Beach, California, to access American capital and NASA launch ranges — but core R&D and engine manufacturing have always stayed in New Zealand.

The Electron rocket made its successful maiden flight in 2017 — a two-stage small liquid-fueled rocket, its first stage powered by nine Rutherford engines (3D-printed, electric-pump-fed LOX/kerosene engines) — purpose-built to deliver small remote-sensing, communications, and defense satellites weighing a few hundred kilograms into low Earth orbit. The Electron is currently the only orbital rocket in the world with an all-carbon-fiber-composite airframe and electric-pump pressurization — its second stage uses a kick stage called Curie for precision orbital insertion.

After 2023, Rocket Lab has successively achieved first-stage parachute recovery and sea retrieval — and has begun developing the larger Neutron rocket, as well as launching its own Photon satellite platform. In other words, it is no longer just "delivering satellites to orbit" — it is directly selling complete mission-data services.

Three things.

Thing one: No university degree — self-taught rocket fuel chemistry and propulsion physics.
Beck is an exceedingly rare kind of "craftsman rocket scientist." The electric pump on the Rutherford engine — battery, motor, pump rotor, all designed in-house — is the topmost branch of a skill tree that began with that old lathe in a New Zealand tin shed.
In your industry — is there a critical technical innovation being invented by someone who did not come through the elite education pipeline — while you are still filtering resumes by degree?

Thing two: The rocket body is not aluminum — it is carbon fiber — cured in a mold inside an oven.
The Electron's airframe is wound from lightweight carbon-fiber composite — not machined from milled aluminum plate. This material process originally came from yacht masts and race-car chassis — Beck brought it into rocket structures, and in doing so drove the dry mass of a small rocket down to the lowest in history.
You — can you take a "mature process from another industry" and transplant it into yours — and turn it instantly into your technical moat?

Thing three: Don't launch the biggest — launch the smallest, the dedicated ones.
SpaceX and ULA fight over massive government contracts and heavy lift. Rocket Lab chose "smallsat express delivery." Others haul dozens of tons of cargo to space in a single wholesale shipment — Rocket Lab sends a few hundred kilos per dedicated courier run.
In your market — is there a niche that "the big players think is too small to bother with" — and is worth your undivided focus?

Peter Beck still walks into the factory every day in a black T-shirt. On his left forearm is a tattoo of a rocket line drawing — the cross-section of the first Rutherford engine, a pencil sketch from 2009 when he was cutting the injector, pass by pass, on a manual lathe in a New Zealand tin shed. He says: "That engine never made it to orbit — but it was the physical proof for Electron."

← Previous Contents Next →
Chapter 182 Cloudflare

2009, San Francisco. Matthew Prince — a graduate of Harvard Business School and Harvard Law School — sat with his former Harvard computer science classmates Lee Holloway and Michelle Zatlyn in a San Francisco bar, trying to solve a DDoS problem that was plaguing a project they had been building for years — Project Honey Pot, an open-source system for tracking web spammers. Every day, Honey Pot was getting hammered by massive volumes of malicious traffic, to the point where their own database was on the verge of collapse. Prince said: "We were protecting other people's websites from spam — but our own system couldn't withstand the attacks." So they wrote a reverse proxy — putting a layer of cache nodes in front of Honey Pot — only letting traffic through to the backend after it had been filtered. That proxy was the seed of Cloudflare. Today Cloudflare is worth roughly 60 billion dollars.

Not CDN. Not DDoS defense. Not edge computing. Not network protocols.
It was "putting a layer of cache in front of Honey Pot" — a defensive shield that four people built to protect their own open-source project — which later became the global reverse-proxy network that protects roughly 20% of all internet traffic.

Matthew Prince was born in 1974 in Salt Lake City. He earned his JD at the University of Chicago Law School? No — at Harvard Law School — and then an MBA from Harvard Business School. He launched Project Honey Pot in 2002 — the goal was to track the bots that crawled websites automatically and dumped spam links into comment sections. They did not go after people; they went after IP addresses. By 2008, Project Honey Pot was being used by tens of thousands of web administrators — but it had itself become a target.

Cloudflare's core is brutally simple: it is a "reverse proxy." Before a user reaches a website, they first connect to a Cloudflare server node at the edge, in some city around the world. That edge node checks the traffic for safety (is it a DDoS? A SQL injection?), accelerates the content (caching images and static files), and only then lets the legitimate request through to the origin server. This idea of "inserting a layer in the middle" — once something only large banks and government websites could afford, using expensive hardware firewalls and dedicated CDNs — Cloudflare turned it into a free plan for everyone. Launched to the public in 2010: small-website protection plus global acceleration — starting at free. Then paid enterprise plans sold unlimited DDoS mitigation, bot management, Argo smart routing, and other advanced features.

Cloudflare went public in 2019. The free-plus-paid freemium model proved extraordinarily hard to disrupt — because millions of free websites feed CF a torrent of attack-signature data — and that data continuously trains the ML models in their global security database, which in turn makes the protection for paying customers even stronger.

Three things.

Thing one: Protecting their own open-source project from attacks — accidentally creating the largest reverse-proxy network on the global internet.
Cloudflare was not a business plan to begin with — it was Project Honey Pot trying to stay alive. This "save yourself first" origin meant they had real attack samples and real attacker IP lists from Day 1.
Did your product come from your own genuine pain point — rather than from a pie chart in a market survey?

Thing two: Free — but the price of free is that all your DNS records pass through CF.
When hundreds of thousands of small websites used Cloudflare's free plan, they entrusted their DNS resolution to CF. DNS is the internet's address book. CF gained visibility into the world's second-largest public DNS network — able to see which country, which AS (Autonomous System), attacks were erupting from.
Do you have a "service that is free — but the free version generates a dataset that no one else can get"?

Thing three: Build "traffic immunity" — not antivirus software.
Kaspersky and McAfee hunt for viruses on endpoints — CF blocks malicious payloads at the "front door" of every website. It does not scan for malware — it bounces malicious traffic away before it even touches you. This is called "edge immunity."

Matthew Prince is still CEO to this day. He likes to publish long essays on his blog and on Twitter — about internet freedom, about country-by-country DDoS statistics, about government censorship and net neutrality. At the end of every product launch, he includes a line he wrote a decade ago: "Protection should be available to everyone — not just to those who can afford to build a fortress."

← Previous Contents Next →
Chapter 183 Air-Products

1940, Detroit. A young chemical engineer from Ohio named Leonard Parker Pool — having earned his PhD in chemical engineering from the University of Michigan — spent several years working in sales and plant management at an industrial gas company (Burdett Oxygen) in Detroit. He noticed one thing: every automobile plant that used oxygen, nitrogen, and acetylene for welding and cutting had to build its own high-pressure cylinder filling station and liquid-oxygen tank farm — consuming enormous amounts of capital and floor space. But if he simply built one large centralized air-separation plant in an industrial park right next to those auto plants — and piped the gas directly to them — each customer could save the millions of dollars it cost to build their own gas station. Today Air Products is worth roughly 70 billion dollars.

Not industrial gases. Not welding. Not chemical engineering.
It was "don't sell the bottle — sell the pipe." Leonard Pool changed the industrial gas business model from "truckloads of steel cylinders delivered one by one" to "a pipe welded onto the customer's factory wall, with a monthly gas bill."

Pool was born in 1903 in a village in Ohio. He grew up near the factories of Detroit — an era when the automobile industry was in the early boom of argon-arc welding and oxy-acetylene cutting. Massive steel plates needed to be cut; molten weld pools needed argon shielding to prevent oxidation. But every factory had to manage its own gas inventory — mountains of high-pressure cylinders, and if they ran out the line stopped — and high-pressure cylinders were hazardous goods, prone to explosion.

In 1940 he built Air Products' first small air-separation plant on the outskirts of Detroit — cooling air to cryogenic temperatures, separating oxygen and nitrogen at their different boiling points — liquid oxygen flowing from the bottom of the distillation column, nitrogen from the top. The liquid products were stored in tanks and then pumped through insulated pipelines to the neighboring auto plant. The customer did not buy equipment — they signed an on-site gas supply agreement and paid a monthly fee per thousand cubic feet of gas.

This pipeline supply model later became the core moat of Air Products — building its own mega-scale air-separation and gasification units beside the giant industrial complexes of the world's steel, chemical, refining, and gasification industries (Jubail in Saudi Arabia, the Changxing Island petrochemical zone in Dalian, China, the U.S. Gulf Coast) — and signing 15-to-20-year "take-or-pay" gas supply contracts with customers, with gas prices indexed to industrial electricity rates and inflation.

In the last decade, Air Products has bet heavily on hydrogen — partnering with ACWA Power and NEOM to build the NEOM green hydrogen plant in the northwest corner of Saudi Arabia (one of the world's largest planned projects for producing hydrogen and ammonia from renewable electricity). At the same time, it is laying out blue hydrogen in Texas, Louisiana, and other regions with carbon-sequestration geological formations — producing hydrogen from natural gas reforming paired with carbon capture and storage. Its core wager is this: the transition over the next twenty years from gray hydrogen to blue hydrogen to green hydrogen — all of it requires air-separation and hydrogen-purification technology — and those technologies are virtually an oligopoly shared by Air Products and Linde.

Three things.

Thing one: Don't charge for the cylinder — charge a monthly "gas rent" for the pipe.
Pool piped gas directly onto the production line — zero upfront investment for the customer. But gas became a bill you had to pay every month, like electricity — and you could not casually switch suppliers — because there was only one pipe welded to your wall.
Can your product shift from "one-time equipment fee" to "pay-as-you-go monthly usage fee"?

Thing two: 15-to-20-year take-or-pay contracts — locking in cash flows that outlive your own lifespan.
The moment Air Products builds a pipeline and invests in an air-separation unit — it signs the customer to an unconditional offtake obligation (pay for the gas volume every month, whether you use it or not) that lasts for decades. This is not gas sales — it is gas infrastructure financing.
How much of your revenue is locked in by "15-year binding" contracts?

Thing three: From oxygen to hydrogen — the underlying physics is the same "cryogenic air separation plus purification" equipment gene.
Air Products' cryogenic separation technology platform — low-temperature distillation, gas purification, high-pressure storage and transport — can extend from air separation to hydrogen production and carbon capture. Three services, with the same physical equipment core at the center — serving downstream markets that can be entirely different.

Leonard Pool retired due to illness and passed away in 1948. But the "don't sell the bottle — sell the pipe" business model he established lives on unchanged in every gas supply agreement of every Air Products global air-separation complex today. His gravestone bears no title — only his name and a chemical formula: "O₂ — N₂ — H₂."

← Previous Contents Next →
Chapter 184 Cintas

1929, Cincinnati. Richard T. Farmer — a former Acme Paper salesman laid off during the Great Depression — scavenged a few hundred returned and slightly damaged industrial rags from Acme's reject warehouse. He washed them, re-dyed them, and drove his own beat-up car around Cincinnati selling these "reborn rags" to factories and repair shops. His pitch was a single line: "Don't buy them. I'll rent them to you — when they're dirty, I come get them, wash them, and bring them back." He named the company Acme Industrial Laundry (later renamed Cintas). Today Cintas is worth roughly 80 billion dollars.

Not laundry. Not chemicals. Not industrial. Not uniforms.
It was "rags are not a product — they are a service."

Farmer was born in Cincinnati around the turn of the century. His first customer was a neighborhood repair shop — the owner complained to him: "Workers just grab the cotton rags to wipe off grease, toss them on the floor when they're dirty — by Friday I count and half the bag is gone." So Farmer stopped selling rags by the pound. Instead, he placed a clean collection bin at the workshop door every week, neatly folded cleaning cotton inside; dirty ones went into a second bin, and on the weekend he collected them and dropped off fresh ones — charging a weekly rental. The factory did not need to buy fabric; it only needed to pay a weekly rental fee for "clean fabric."

In the 1940s and 50s — Cintas expanded from industrial rags to in-house employee uniforms — workers' coveralls, shirts, anti-static garments, white coats for food processing. Farmer's pricing logic stayed the same — the factory did not buy the clothes; it paid a weekly per-person "uniform cleaning and replacement" service fee. Cintas made a tiny profit on every uniform it cleaned and returned intact each week. Millions of uniforms — collected, washed, pressed, folded, and delivered back every week — billed per week, per piece. Tiny profit times millions equals vast.

Then it expanded laterally — from uniforms to floor mats (entrance anti-slip mats, logo mats — replaced and cleaned weekly), to restroom supply dispensers (paper towels, hand soap, air fresheners — billed per "single use"), to first-aid and compliance training services (OSHA worker safety training, first-aid kit replenishment — renting a safety compliance scorecard), to fire extinguisher and equipment safety inspections — every one of these added a "weekly service touchpoint."

Three things.

Thing one: Rags are not for sale — they are for rent. Dirty ones swapped for clean, free of charge.
From "buy cotton rags once" to "rent clean rags weekly," Farmer offloaded the factory's washing labor onto his own laundry plant — and the factory was happy to pay a weekly rental to never have to manage cleaning itself.
Can your product shift from "buy it and own it" to "rent it, and we replace it on schedule"?

Thing two: Touch the customer every week — increase your service density.
The most valuable thing Cintas owns is not the rags — it is "that Cintas driver who shows up on time every Tuesday morning at the workshop door, drops off clean uniforms, picks up the used ones." This fixed weekly appearance turns any new service need into an "add one more thing while I'm here" — at an extraordinarily cheap incremental cost.
Do you have a "weekly knock on the door" practice — that lets you keep adding new services without friction?

Thing three: The competition is not about uniforms — it is about laundry capacity.
Cintas built its own massive industrial laundries — high-temperature, sanitizing, sorting, pressing, packaging assembly lines. These laundry assets and logistics routes are a physical-infrastructure competitive moat that cannot be easily replicated.

Richard Farmer retired in the 1970s and passed away in his later years. His grandson Scott Farmer took over as CEO in 2015 — turning Cintas into a Fortune 500 company. Cintas's American cleaning-and-logistics network, operating at the trivial scale of weekly paper-towel replacement and fire-safety inspections, has produced shareholder returns that outperform most technology stocks.

← Previous Contents Next →
Chapter 185 TransDigm

1993, Cleveland. A finance and M&A lawyer named Wally Nichols — who had spent years in venture capital and private equity — led a small team in buying the aerospace brake components division of Kelsey-Hayes. Kelsey-Hayes was a manufacturer of aircraft brake actuators and hydraulic components that its giant conglomerate parent had forgotten about. But Nichols had discovered a profit model that nearly every aerospace supplier overlooked: once an aircraft model goes out of production — for the remaining 30 to 50 years of that aircraft's flying life — the market for original replacement parts (OEM spares) is a fully sealed, exclusive monopoly. Today TransDigm is worth roughly 80 billion dollars.

Not aerospace. Not engineering. Not intellectual property.
It was "locking in the exclusive supply rights for replacement parts on out-of-production aircraft." Once an aircraft (say, the Boeing 737 NG) stops being built — all replacement parts for the airframe structure, propulsion, landing gear, and flight controls — only the original certification-holding supplier can produce them — the FAA prohibits any new, uncertified manufacturer from entering. Because the cost of re-certifying a single landing-gear spare part — the testing alone exceeds all the revenue that part could ever generate from sales. TransDigm's business is to buy these orphaned spare-parts businesses from giants like GE, Boeing, Airbus, and Honeywell — lines those giants consider "too small to bother with" — and then supply them at the highest possible margins.

Nichols was born in the 1950s in Cleveland. He studied law at Cleveland State University, practiced corporate and M&A law locally, then managed investments at several private equity funds. He specialized in hunting down "the forgotten little aerospace parts divisions inside big conglomerates." When he found a small business unit buried inside some aerospace systems contractor — something with only 100 to 200 million dollars in annual sales but extraordinarily high margins and customers who could not switch suppliers — he would buy it from the parent company at an extremely low multiple.

TransDigm's customers are airlines and air force maintenance depots. When an aircraft has a faulty component — the maintenance engineer must source a replacement part by its OEM part number — and FAA airworthiness directives ban the use of uncertified parts. If TransDigm is the sole OEM certification holder for that specific part number — the customer has exactly two options: pay TransDigm's asking price — or ground the aircraft.

The moral controversy around this business model is enormous — members of Congress have repeatedly summoned Nichols to testify on whether he is "excessively overcharging the military." His reply: "What we sell is — the avoidance of a 200-million-dollar aircraft being grounded for want of a 20,000-dollar valve."

Three things.

Thing one: Buy the part that a giant considers "too small to bother with" — but for that part, you are the entire market.
Every TransDigm acquisition does not chase market share — it chases that single FAA parts authorization as the world's sole certified manufacturer. When you are the only one — the price is not set by the market.
In your industry — is there a "small" domain that a giant abandoned because it was too small — but once you acquire it, you become the sole global supplier?

Thing two: It is not proprietary technology — it is proprietary FAA certification.
The moat is not patents — it is "FAA part number registration." If you already hold that number — no one else can follow. The cost of entry exceeds the total addressable market.
Where is your legal or regulatory moat?

Thing three: Don't sell new parts. Don't grow. Don't do R&D. Just sell the "lifetime guarantee" on out-of-production parts.
Most of TransDigm's products are for aircraft that are no longer in production. They do not sell to incremental growth — they hold the installed base — because the gross margin on a single replacement valve can reach 80% to 90%.

Wally Nichols stepped down as CEO in 2018. In interviews he rarely discusses business ethics. When asked, "How does it feel to sell a valve that costs a few thousand to make but, once you acquire it, becomes a spare part priced in the hundreds of thousands?" — he said: "Airlines don't call to complain — because they're buying a valve, but they're saving the millions it costs to ground a plane. If you wrote the valve price on the other side of a piece of paper labeled 'one day of grounded aircraft' — you might think my price isn't high enough."

← Previous Contents Next →
Chapter 186 Motorola-Solutions

1928, Chicago. Paul V. Galvin — a former battery salesman who had lost everything when his company went bankrupt — rented a small workshop on West Harrison Street in downtown Chicago. He and five employees began assembling battery eliminators and radios into large wooden cabinets on a conveyor belt. Galvin hung a sign in the window that read "Galvin Manufacturing Corporation." He named his first mass-produced car radio the Motorola — "motor" plus "ola," meaning "sound in motion." When the first Motorola car radio was installed in a Studebaker in 1930, an entirely new category was born in the automobile industry. Today Motorola Solutions is worth roughly 60 billion dollars.

Not mobile phones. Not radio. Not public safety. Not the FCC.
It was the impulse to "cram a radio into the dashboard of a car that still runs." Galvin had heard it — passengers on long drives needed more than the wind outside the window.

Paul Galvin was born in 1895 in Harvard, Illinois. He briefly attended the University of Illinois, served as a signalman on the Western Front in World War I, then returned home and failed at two businesses. In 1928 he decided to "open a third company" in the same workshop — by then he and his partner were so broke that parts suppliers refused to ship them materials. He took the only thing of value he had left — a gold watch — and pawned it to pay his workers' wages.

The first Motorola radio was demonstrated at the 1930 National Radio Exhibition — and drew several hundred orders on the spot. It was nothing like a modern embedded radio — it was a large standalone metal box that stretched across the lower edge of the dashboard. Installation required drilling through the firewall to run the wires. But after that, Galvin took Motorola radios into police cruisers (two-way radios) and into the military (during World War II, Motorola built the first portable battery-powered infantry radio — the SCR-300 — weighing about 32 to 38 pounds, strapped to a radioman's back so he could talk while marching). That was the ancestor of every "walkie-talkie" in the world.

After the war, Motorola pioneered the mobile phone — in 1973, Martin Cooper placed the world's first portable cellular phone call on a Motorola DynaTAC prototype while walking down Sixth Avenue in New York. But in 2011, Motorola was split in two: Motorola Mobility (consumer phones) was sold to Google and later handed off to Lenovo — while Motorola Solutions retained all public-safety, police, fire, and emergency-services private-network two-way radios, command centers, and surveillance/drone operations.

Today's Motorola Solutions is the largest supplier of public-safety communications systems on earth — more than 90% of police departments and emergency services in the United States use its proprietary LTE and two-way radio systems — as do their counterparts in London, Hong Kong, Singapore, and Australia. The core is a closed-loop ecosystem called FirstNet, the APX handheld radio, and command-and-control centers. People do not think about the brand — but every time a police officer presses the push-to-talk button on their shoulder radio, the unit on the other end is very likely a Motorola.

Three things.

Thing one: When completely bankrupt — pawn the last thing of value to pay labor — then launch the world's first car radio.
The day Galvin pawned his watch — he stood at the factory door and told his five workers: "If this radio can play three full songs in a real car — we make it."
Can your willpower — when everything is gone — still give a new product one shot?

Thing two: Make police, fire, and military signaling independent of the civilian cell network.
When mobile networks get jammed in a crowd — public-safety networks need dedicated frequencies and 100% reliability. Motorola Solutions built a private network that exists only for "emergency" — and that "never congested" guarantee is the moat.
Are you competing on an "open platform" — when you could build a "dedicated channel"?

Thing three: Split yourself apart — send consumer electronics into another body.
After the 2011 split, Motorola Solutions shed the mobile phone — the headline brand — and focused on building irreplaceable public-safety infrastructure. The consumer phone shell died — but the lifeline system lives on.

Paul Galvin passed away in 1959. Throughout his life, his office held a piece of glass from an old Studebaker — embedded in the dashboard behind it was the prototype of the first Motorola radio in the world, long since unable to power on. But every young engineer who came in for an interview had to look at it before they sat down. In his memoir he wrote: "Business is good — if someone else needs you."

← Previous Contents Next →
Chapter 187 MicroStrategy

1989, Wilmington, Delaware. An MIT engineering graduate who had spent a few years doing biostatistics and numerical modeling at DuPont — Michael Saylor — founded a company together with his MIT fraternity brother Sanju Bansal. The company's earliest business was not "buying Bitcoin" — it was business intelligence (BI) software: pulling corporate databases, sales records, inventory, and financial data into a platform called MicroStrategy and turning them into draggable, interactive charts. During the dot-com crash in 2000, MicroStrategy was investigated by the SEC over its revenue-recognition accounting and its stock dropped 62% in a single day. But in August 2020, Saylor made what may be the most counterintuitive corporate finance decision of the 21st century — he converted all of the company's cash into Bitcoin. Today MicroStrategy (now Strategy) is the publicly traded company holding the most Bitcoin in the world — worth roughly over 50 billion dollars.

Not cryptocurrency. Not blockchain. Not Web3.
It was Michael Saylor using his MIT mathematical training — calculating that "the purchasing power of the U.S. dollar declines X percent every year — while Bitcoin has an absolutely scarce supply curve" — and then converting every dollar of his company's excess cash, plus the money raised from debt issuance, into Bitcoin.

Saylor was born in 1965 in Nebraska. At MIT he earned a dual degree in Aeronautics and Astronautics and Science, Technology, and Society. In 1989 he founded MicroStrategy — one of the pioneers of the BI field. BI is dull but extremely stable B2B software — the customers are large corporations, banks, and governments. This business gave MicroStrategy a steady but slow, semi-stable cash flow for over two decades — but it never took off.

In March 2020, amid the COVID crash, Saylor began to read and study the Bitcoin white paper and monetary theory, line by line. In August 2020, he announced that MicroStrategy had purchased 250 million dollars' worth of Bitcoin as its "primary treasury reserve asset." All of Wall Street said the man had lost his mind — the balance sheet of a software company, fully allocated to Bitcoin.

Between 2020 and 2025, MicroStrategy — through multiple issuances of convertible bonds and stock sales — poured a cumulative total of over 15 billion dollars into Bitcoin. The price of Bitcoin went from roughly 10,000 dollars in 2020 to well over 100,000 dollars by early 2025. The company's stock, MSTR, has essentially become a "leveraged Bitcoin ETF" — its share price highly resonant with the Bitcoin price, with staggering volatility — attracting two entirely different groups of investors: crypto狂热者 and traditional hedge-fund arbitrageurs.

By 2025, Strategy (the company dropped the "Micro" prefix) holds over 500,000 Bitcoin on its balance sheet. Saylor himself posts Bitcoin-related technical and philosophical commentary on X every day — citing everyone from Newton and Shannon to Hayek on information entropy and monetary theory — and has declared: "Bitcoin is not a currency — it is a form of digital energy stored in cyberspace."

Three things.

Thing one: A C-suite executive — using math to calculate reserve assets — and then going all in.
Saylor is not a "crypto believer" — he used the mathematical methods of his MIT engineering training to calculate the currency-devaluation curve and Bitcoin's deterministic supply curve — and then executed the conclusion in full.
Have you calculated what percentage of its purchasing power your cash holdings have lost over the last decade — and then looked again at your cash-reserve strategy?

Thing two: Turn the company into a "Bitcoin bridge" — issue convertible bonds to buy coins.
MicroStrategy used its status as a public company — issuing bonds at low interest rates or selling new shares — and converting those proceeds into Bitcoin — effectively giving shareholders an "amplifier" on Bitcoin's price appreciation.
Could you — using your corporate structure — perform a kind of "arbitrage" for your customers or shareholders that only you can do?

Thing three: A software company — whose market cap is no longer determined by the valuation of its software business.
MicroStrategy's stock price today is almost entirely determined by the price of Bitcoin — its BI software business has become negligible. The BI business is simply the legal wrapper that allows MSTR to function, in essence, as an "SEC-regulated Bitcoin trust fund."

Michael Saylor temporarily stepped down as CEO in 2022 following an insider-trading settlement with the SEC, but remained Executive Chairman. He still posts over a dozen Bitcoin tweets every day. He once said: "When I was 25 as an MIT student — I thought I would invent cold fusion. At 55 — I discovered that the only already-invented form of absolutely scarce energy is Bitcoin — and what I can do is buy it — and tell others why."

← Previous Contents Next →
Chapter 188 The-Travelers-Companies

1864, Hartford, Connecticut. James G. Batterson — a former stonecutter and railway ticket agent — hung a new sign in his downtown Hartford office. Two months earlier, near the Bank of England, he had encountered the phrase "accident insurance" for the first time — the idea that accidental injuries a person suffers in everyday life, outside of work, could also be insured and compensated. At that moment, not a single company in the United States offered such a policy. Batterson returned to Connecticut and founded the first Travelers Insurance Company — covering only two things: accidental injury while traveling, and accidental death. The premiums were extremely low, the policy amounts small, but this was the first time in human history that insurance was extended from "the ship sank" and "the house burned down" to "a living person fell off a carriage on the way to the post office." Today The Travelers Companies is one of the largest property-casualty insurers in the United States, worth roughly 60 billion dollars.

Not auto insurance. Not home insurance. Not reinsurance. Not actuarial science.
It was "a person fell off a carriage" — an everyday risk that every insurer had ignored, and he turned it into a new class of coverage.

Batterson was born in England in 1820 and immigrated to America as a teenager. He worked as a stonecutter in Connecticut and then as a railway agent — and through his contact with railway passengers and workers, he saw again and again how a minor injury or a single broken bone could plunge a family into destitution overnight — because no insurance covered it. The only insurance that existed at the time covered massive marine cargo and fire. In 1864 he raised less than 5,000 dollars and opened the first Travelers Accident Insurance Company — his first customer was himself.

The terms of his first-generation accident policy were brutally simple: if a traveler died accidentally while traveling by "public conveyance" — the payout was 5,000 dollars. The annual premium was 5 cents. Today Travelers' policies span everything from auto insurance to comprehensive general commercial liability, workers' compensation, cyber insurance, and D&O liability — but its origin — "insuring a worker on his way home from the factory" — still lives on in every workers' compensation policy today.

In the 1890s, Travelers introduced early automobile liability insurance. It then expanded into employer liability, commercial property, and surety — and since 2016 has been one of the largest independent property-casualty insurers in America. It remains, to this day, the only insurance company included in the Dow Jones Industrial Average.

Three things.

Thing one: See a risk that has never been "insurance-ified" — and build the first category.
Batterson went to England, saw "accident insurance," transplanted it to America, and turned it into a product called Travelers. Travel risk — was the very first subject matter for personal-lines insurance.
Are there behavioral risks in your daily exposure — risks that have never been commercially insured or protected?

Thing two: Start with penny premiums — accumulate them a millionfold.
Early accident insurance premiums were tiny — but the accumulated claims data gave Travelers, earlier than any later entrant, knowledge of accident frequency across different industries and modes of travel — and therefore the ability to price risk more accurately.
Was your data accumulated through a "massive volume of extremely low-priced transactions"?

Thing three: An insurance company's name — can itself build brand trust in reverse.
Most insurance company names were things like "Hardware Mutual" or "Fireman's Fund" — people were supposed to entrust their money to an abstract corporate entity. But "The Travelers," with its red umbrella logo — planted the image directly in people's minds: "You will be protected on your journey."

James Batterson died in 1901. On the rooftop of Travelers' headquarters building in Hartford, a giant steel sculpture of a red umbrella still stands to this day. Sometime in the early 1900s, a manager had someone bring an actual umbrella to the roof and open it up — and the umbrella has remained there ever since. What it represents is not "shelter from rain" — it is "someone holding an umbrella open above your head, perfectly positioned to block the bad luck that blindsides you on a clear sunny day."

← Previous Contents Next →
Chapter 189 Baker-Hughes

1907, Navasota, a small oil town on the Texas Gulf Coast. A mechanic named Reuben C. Baker — who had been maintaining steam engines and mud pumps for local drilling contractors — invented an extraordinarily simple little tool in his small workshop: a "casing annulus cement retainer" that could be lowered into the wellbore on a cable. Before Baker, if you wanted to pump cement into the space between the wellbore and the steel casing to prevent groundwater seepage — you had to pour cement into that annular gap by hand — and the success rate was abysmally low. Baker's cement retainer was an expandable rubber cylinder: lowered to the intended position, it used hydraulic pressure to push the rubber outward, compressing it into the annular space to form a complete seal — and then cement was pumped downward on top of that seal. For the first time, this tool allowed an oil well to "reliably isolate underground aquifers from the production channel." Today Baker Hughes is one of the world's three largest oilfield services companies, worth roughly 40 billion dollars.

Not oil. Not drilling. Not hydraulic fracturing. Not geology.
It was a rubber sleeve that could be expanded outward by hydraulic pressure, pressing tight between the steel pipe and the rock.

Reuben Baker was born in 1873 in Oregon and spent his youth working as a mechanic and tool salesman across the western oil fields. He noticed that the most common and dangerous drilling accident — the blowout — was almost always caused by groundwater breaking into the wellbore. The physical solution was to inject cement into the annular gap between the well casing and the drilled formation — but that gap was an annular cavity thousands of feet underground, and no ready-made sealing method existed. Baker combined rubber and a hydraulic cylinder — and built the first expandable retainer that could truly seal off the annulus deep within the formation.

He founded Baker Tools in 1907, gradually expanding the product line to drill bits, downhole measurement tools, and mud pumps — folded into the full "drill, measure, complete, produce" workflow of oilfield services. In 1987, Baker Tools merged with Hughes Tool Company (which had grown out of the early enterprises of Howard Hughes Sr. — his twin-cone rotary roller bit had opened up deep-well drilling to the whole world in 1908) — creating Baker Hughes.

During the shale revolution of the 2010s, Baker Hughes was the target of an attempted acquisition by Halliburton (later blocked by regulators) — and in 2019 a majority stake was sold to General Electric, forming GE Baker Hughes. Then, between 2021 and 2023, GE gradually reduced and exited its stake, and Baker Hughes once again became a fully independent international oilfield services company. It now holds a significant position in rotary steerable drilling systems, compressors, gas-turbine compression, and LNG liquefaction processes — and it uses its oilfield-services cash flow to invest in geothermal energy, carbon capture and storage (CCS), and clean-hydrogen production technology.

Three things.

Thing one: A rubber sleeve — that sealed the well — the "safety foundation" of all modern drilling.
Without the Baker retainer — deep oil and gas wells cannot reliably isolate hundreds of underground aquifer layers. It becomes nearly impossible to prevent catastrophic shallow gas from migrating to the surface. A simple rubber-and-hydraulic physics design — underpins the wellbore integrity of millions of deep wells around the world.
Is your product the most fundamental "safety component" rather than the most expensive "main component"?

Thing two: Merging with the greatest drilling invention (the Hughes bit) — assembling the tool chain from surface to subsurface.
Baker (completion + packers + fluids) and Hughes (bits + drilling parameters + mud) integrated — so an oil company could receive a complete "wellbore construction service package," from spud to completion, from a single supplier.
Could you merge with a company that is perfectly complementary to you but has zero overlap — to become a one-stop terminal?

Thing three: From oil to geothermal to carbon storage — wellbore technology translates.
Baker Hughes' drilling and completion technology — suited for extracting oil — is equally suited for sequestering CO₂ deep in the earth and developing geothermal energy. Its rotary steerable and drill-bit systems can drill into geothermal granite basement — the same rig.

After Reuben Baker passed away in the 1950s, that first workshop in Navasota was preserved and remains a small oil museum to this day. Inside there is only an old lathe, a few cracked rubber molds, and a yellowed workshop notepad — on which he drew the first pencil sketch of the retainer and wrote a single line: "The hole must not leak."

← Previous Contents Next →
Chapter 190 Republic-Services

The 1980s, Phoenix, Arizona. Wayne Huizenga — a serial entrepreneur who would later create Blockbuster Video and Waste Management — and his associates merged a string of independent local waste-hauling companies into a chain called Republic Waste Services — and then kept acquiring more local collection routes. Later, Republic Industries (Huizenga's mothership) focused on autos and entertainment — and this waste-asset package was spun off in 1996 as the independent Republic Services. Today Republic Services is the second-largest solid-waste management and recycling company in the United States, behind only WM (Waste Management), worth roughly 60 billion dollars.

Not environmentalism. Not recycling. Not zero waste. Not circular economy.
It was "monopolize a municipal or county-level garbage collection route, and then let the cash flow from that route grow at a steady, identical rate year after year — because no one competes with you, since it is impossible for a second company to run two garbage-truck fleets on the same street."

The core asset of Republic Services is not the recycling plant — it is "the route." Every residential collection route, every commercial waste compactor — once a long-term contract is signed and the waste stream is established with the local city or county government — it is almost eternally irreplaceable. Because the physical density of truck routes means the cost of "adding one more household" approaches zero — while building an entire collection and transfer-station network from scratch is prohibitively expensive for any new entrant.

In 2008, Republic executed a merger that shook the entire industry — merging with Allied Waste to form a unified solid-waste and recycling network covering over 40 U.S. states. After that, its main growth came from acquiring small independent private waste haulers — many of them family-run. Republic then gradually built its own modern material recovery facilities (MRFs) and landfills (with LFG collection and power generation), turning "landfill gas" into another profit center.

In recent years, Republic has taken the lead in the U.S. waste industry in large-scale piloting of electric garbage trucks and automated side-arm collection routes — using GPS and fill-rate sensors to optimize routes and reduce empty runs. At the same time, it has expanded its "environmental solutions" — including industrial hazardous waste treatment, drilling-waste solidification and landfilling, and on-site emergency spill response.

Three things.

Thing one: Route density — that is the waste industry's "network effect."
The more neighbors you have on the same street — the lower the average cost per stop for garbage collection. Each Republic route can service 100 to 150 bins per hour — marginal cost pushed to the absolute floor.
Does your business have a natural monopoly effect created by "physical density"?

Thing two: Merge — but weave the routes into an "irreplaceable" patchwork quilt.
When Republic merged with Allied Waste — the merger did not generate serious antitrust problems — because the combined routes could be reorganized across adjacent markets to eliminate overlap and further reduce the cost per ton of waste.
In your merger — can you "assemble a nine-square grid" on the physical map — turning overlap into a more efficient contiguous whole?

Thing three: Landfill gas — extracting residual value from the waste stream.
Beneath the cover layer of Republic's landfills, gas-extraction wells are inserted — methane is collected and used to generate electricity or converted on-site into compressed natural gas — fueling the garbage truck fleet itself. It is, in effect, pulling a second layer of "fuel revenue" out of the buried waste.

Wayne Huizenga passed away in 2018 — the three entirely unrelated companies he personally created (Blockbuster, Waste Management, AutoNation) each became the largest player in its respective industry. Republic Services traces part of its lineage to the remnants of Huizenga's early waste ventures. Republic's headquarters is now in Phoenix — and every morning in that desert city, thousands of white-and-green garbage trucks roll out, their radios possibly playing some old-school rock that no one can quite name.

← Previous Contents Next →
Chapter 191 Nike

1964. A Stanford MBA graduate and college track runner borrowed 50 dollars from his father and drove his green Plymouth Valiant to every high school track meet in Oregon. In his trunk were Japanese running shoes — Onitsuka Tigers. After the races, he worked the crowd like an ice-cream vendor, pitching coaches and athletes one by one. "Try these — lighter than Adidas." Today, Nike is worth 150 billion dollars. What crawled out of that green trunk is the most valuable sports brand on earth.

Not inventing the running shoe. Not a technological breakthrough. Not sports science. Not celebrity endorsement.
It was Phil Knight. An Oregonian with a natural stammer who blushed when speaking in public. In four years of college, he never won a single important race on the middle-distance track. The greatest gift of his life was not running fast — it was that he understood, better than anyone on earth, "what the runner feels underfoot on the road at five in the morning." And with that understanding, plus a "swoosh" logo he bought for 35 dollars from a Portland State University student, he beat Adidas and every sports brand bigger than his own.
Let me tell his story first. When I'm done, we'll talk about why "knowing what the bottom of the user's foot feels" is worth more than any market research.

Phil Knight's track coach when he was an undergraduate at the University of Oregon was Bill Bowerman — one of the most legendary track coaches in American history, and later the co-founder of Nike. Bowerman's most famous habit: on Sunday mornings he would go to the rubber track in Eugene, personally take apart his athletes' spikes, and re-stitch the soles using hot rubber from his home oven and new nylon materials. He was not a cobbler — he was an obsessive who "wanted to shave off the last ounce of shoe weight." Every shoe he ever took apart — Adidas, Onitsuka Tiger — he improved.

Knight wrote a paper at Stanford Business School: "Can Japanese running shoes defeat German athletic shoes the same way Japanese cameras defeated German cameras?" His professor gave it a B. Said it "lacked data to support the thesis."

1962 — Knight was 24. After graduating, he bought a round-the-world plane ticket. Not for a vacation — he flew to Kobe, Japan, and walked into the offices of Onitsuka Tiger. He told the manager of the shoe company: "My name is Phil Knight, and I represent a company on the West Coast of the United States called Blue Ribbon Sports." — There was no such company. He made it up on the spot. He delivered an impromptu plan for expanding Japanese shoe sales in the West Coast track market. Onitsuka Tiger believed him. Gave him the distribution rights.

Back in Oregon, he pulled the first shipment of 12 pairs of Tiger running shoes from his trunk and stashed them in Bowerman's garage. The two each put in 500 dollars and founded Blue Ribbon Sports — the predecessor of Nike. Then Knight started standing on the sidelines at every high school track meet — trunk open, shoeboxes stacked in a single layer. Every time a coach walked past, Knight would watch the race with him, talk about the weather, talk about some kid's problem with the final ten meters of the sprint — and then casually mention that he had a pair of Japanese shoes in his trunk, lighter than what the coach was wearing. "Try them — no charge. Pay for the second pair."

In 1971, their relationship with Onitsuka Tiger fell apart. The distribution contract was unilaterally terminated — the equivalent of "someone cutting off your supply line." Knight had to find a replacement source and build a new brand in an instant — and he had no money for advertising.
He flew to Mexico to inspect a factory, then flew to Japan again to partner with a new contract manufacturer. Then he approached a graphic design student at Portland State University named Carolyn Davidson. He gave her 35 dollars — roughly the price of a textbook in that era — and asked her to draw a logo that "looked like speed, like a foot in motion, like a gust of wind." When she turned in the work, she herself said it "didn't really look like anything specific." Knight glanced at it and said: "I didn't like it at first. But I'll learn to like it."

That was the Swoosh. Years later, Knight gave Davidson much more — significant stock and a custom gold Swoosh ring — but in 1971, all he could pay was 35 dollars.

Then Knight and Bowerman made a shoe. They stole Mrs. Bowerman's Belgian waffle iron, poured liquid rubber into the grid of the iron — trying to create a "honeycomb sole that increased grip." The iron exploded. The rubber scorched onto the stovetop. Mrs. Bowerman came home, smelled the burning, investigated the kitchen, and found her waffle iron had become a charred, irregular slab of rubber. They scavenged through the wreckage, pulled out a small reasonably uniform section of grid-pattern sole — and test-built one shoe.
That was the Waffle Trainer — the honeycomb waffle sole. Its grip on wet concrete outperformed anything Adidas or Puma had at the time by a clear margin. And it was dirt cheap — a waffle-iron mold — the production cost was extremely low.
At the 1972 Munich Olympics, American marathon runners wore Nike for the first time in a major international event. In 1984, they signed Michael Jordan.
The story after that, the whole world knows — Air Jordan, Just Do It, Tiger Woods, LeBron, Serena Williams.
But here is an important detail: Nike never invents running shoes. It has never been "the world's leading cushioning-technology lab." Its core methodology is — "repeatedly test and refine with the very best athletes, then attach their names to the product." Athletes build the shoe, not the lab. It pays the highest endorsement fees in the world — and then those athletes' feedback is directly converted into physical design revisions for the next generation of product. And then the consumer puts on "that person's shoe," not "a technology."

Phil Knight retired in 2016. He wrote a memoir called Shoe Dog — at the end he wrote: "The first thought I wake up with every morning is still the same — can that shoe be one ounce lighter. When that young runner steps onto the track at five in the morning, in minus-five-degree weather, will he feel like someone is running beside him."
He is 84 years old. He says he is still thinking about that shoe.

Three things.

Thing one: Be your user's "kin," not their "supplier."
Knight himself was that runner doing 5 a.m. miles. He sold running shoes to make that early-morning road a little easier. Every one of his marketing decisions came from "what the runner's foot feels on the road tomorrow" — not from "this quarter's footwear sales trend report."
Do you have an advantage — that you yourself are the extreme version of your target user?

Thing two: What you sign is not a star — it is an "embodiment of faith."
Nike picked Jordan not because he played well. It was because every teenager in the world wanted to become Jordan. And then the Nike shoe became the first tangible prop for that "becoming."
Do the spokespeople and customer stories you choose — make someone want to become that person the second they see them?

Thing three: Always be close to going under.
From Day 1 to retirement, the phrase that never left Knight's mouth was — "We have six months until we're bankrupt." Even when Nike's market cap surpassed 100 billion dollars. He demanded that the CFO always report the numbers at "maximum crash speed." This made Nike the fastest-reacting company among its peers — because it pretended it was always the rabbit being chased.
How safe do you feel right now? Has that safety already made you as slow as a giant?

That 35-dollar Swoosh is now painted on every basketball court on every continent. And Phil Knight says he doesn't look at the logo — he still looks at the waffle grid on the sole.

← Previous Contents Next →
Chapter 192 Keysight

1939, Palo Alto, California — right beside Stanford University. Bill Hewlett and Dave Packard baked their first audio oscillator — model HP200A — using a second-hand Sears electric oven in a garage, and sold it to Walt Disney's animation studio to test the frequency response of the surround-sound system for Fantasia. That garage has been officially marked by the State of California as "the birthplace of Silicon Valley." In 1999, Hewlett-Packard spun off its entire test-instrument, life-science, and analytical-instrument businesses — creating Agilent Technologies. In 2014, Agilent spun off its pure electronic test and measurement business yet again — creating Keysight Technologies. Today Keysight is worth roughly 40 billion dollars, and it owns the world's fastest real-time oscilloscope and the widest-bandwidth vector signal analyzer.

Not electronic information. Not communications protocols. Not 5G. Not semiconductors.
It was "measurement." Using extraordinarily precise instruments to measure a signal's voltage, frequency, phase, and timing — and from those few physical values, determining whether an entire system can pass electromagnetic compatibility, whether it can demodulate correctly above 6 GHz. This layer of "physical-layer verification" — as human communications advanced from 1G to 2G to 3G to 4G to 5G to 6G — has run continuously on the same instrument product line that stretches from HP to Agilent to Keysight.

Dave Packard and Bill Hewlett founded HP in 1939 in the garage at 367 Addison Avenue in Palo Alto. Early HP made only electronic measurement instruments — frequency counters for signal sources, voltmeters, oscilloscopes. These were basic tools for the R&D and repair of electronic equipment. In this dull but precision-critical category, HP gradually came to monopolize the test benches of the world's most advanced laboratories. The birth of the semiconductor, satellite communications, the rollout of internet switching equipment — not a single one left the factory without first being debugged with HP test equipment.

In 1999, HP carved out its instruments, medical, and chemical-analysis businesses as Agilent. In 2014, Agilent transferred all electronic instruments into Keysight. As an independent company, Keysight has continued to push the measurement boundaries of millimeter-wave, wideband, and ultra-low phase noise — and has become the irreplaceable "eye-diagram" instrument supplier for fields like the SerDes (high-speed serial transceivers) in AI training chips, PCIe 6.0, and DDR5 memory compliance testing.

Three things.

Thing one: From the surround sound of a Disney theater to the 224G SerDes eye diagram inside an AI training server — the same physical essence.
It is still "measuring a voltage change over time" — only the frequency has gone from hundreds of hertz to 224 GHz. The measurement physics Keysight inherited remains unchanged — the instrument bandwidth simply marches to higher frequencies as time goes on.
Do you have a fundamental physical measurement capability — that can persist across industries for a hundred years?

Thing two: Every time communication technology advances a generation — the test tools must be one generation ahead.
When a chip company designs a 5G modem — Keysight must, two years earlier, already be producing instruments that can transmit and analyze 5G NR signals. What it does, always, is build the test environment for the "next-generation communication that still has no chip."
Are you helping others "build the future" — while yourself standing one step further ahead, in the "pre-future"?

Thing three: Spin off and spin off again — leaving the instruments with the least brand recognition but the most indispensability for last.
HP to Agilent to Keysight — each divestiture retained the instruments business, which carried lower margins but extraordinarily deep customer stickiness. What remained at the end was the pure scientific measurement core — with the highest technological moat, the lowest brand recognition, and the deepest customer lock-in.

Keysight's headquarters is in Santa Rosa, California. The company name Keysight combines "Key" and "insight" — meaning "gaining insight through critical measurement." In the lobby of the headquarters sits a restored HP200A audio oscillator — beside it a plaque that reads: "This device measured the surround sound for Fantasia. Its descendants are now measuring the background noise of the universe."

← Previous Contents Next →
Chapter 193 Sempra

1998, San Diego, California. Amid the great wave of deregulation in the California electricity market, a local natural-gas utility called Enova — whose roots traced back to the San Diego Gas Light Company, founded in 1881 — merged with Southern California Gas Company (Pacific Enterprises) to form Sempra Energy. The new brand name, conjured from scratch in the late 1980s, drew from the Latin word semper, meaning "always." Today Sempra is worth roughly 50 billion dollars.

Not power trading. Not retail energy. Not shale. Not solar PV.
It was "building a transmission and distribution network that straddles both electricity and natural gas, in the southwestern corner of the North American continent — near the Mexican border, near the Pacific Ocean, near an LNG receiving terminal."

What makes Sempra unique — unlike most American utilities that merely distribute electricity or pipe gas within a single state — is that it crossed the border into Mexico. It built natural-gas pipelines connecting Texas to northwestern Mexico; it built and operates large wind farms in Mexico, selling power to local industrial users; and at the same time, it has built and operates LNG export terminals on the U.S. West Coast — at Cameron LNG in Louisiana and ECA LNG in Ensenada, Baja California.

Its underlying logic is this: California itself provides regulated electricity and gas revenue that is stable but low growth — but "running a pipeline across the border" and "connecting the Southwest U.S. pipeline network to Asian LNG buyers" generates rate streams with much higher growth. This strategy turned a "utility" into an "international energy corridor."

Three things.

Thing one: Don't generate power — provide "transport + storage + regasification" — when global natural gas demand is highly dispersed across upstream wellheads, and the consumers are on the other side of the Pacific.
Sempra does not bet on the price of natural gas — it builds a corridor from the Permian Basin to a Mexican LNG export terminal. A cross-border pipeline equals geophysical lock-in — revenue is secured by multi-decade take-or-pay contracts.
Are your assets the "physical connector" between two national markets?

Thing two: The name means "always" — because infrastructure is pipe — with an expected depreciation life of over 50 years.
What Sempra builds is not a digital app — it is steel pipeline and liquefied-gas storage tanks — designed to still be operational in 2060 to 2070.
When you build your assets — do you consider whether "people 30 to 50 years from now will still need them in the same physical form"?

Thing three: Hold the natural-gas grid plus renewable generation on both sides of the border — and tie them together with LNG.
California's grid has a surplus of solar during the day — Sempra can use its gas pipeline network to inject hydrogen, made from surplus renewable electricity, into the natural-gas pipeline system (in the future). What it is doing is fusing three networks into one: gas, electricity, and hydrogen.

Sempra's current CEO, Jeffrey Martin, has been at the helm since 2018. He often tells his executive team: "Pipelines and power lines are not going away — they will be 'reclaimed' — carrying methane today, carrying hydrogen tomorrow — but the physical channel is the same pipe."

← Previous Contents Next →
Chapter 194 Carvana

2012, Phoenix. Ernie Garcia III — a young man only a few years out of Stanford Law School — had been working in sales operations at his father Ernie Garcia II's used-car auction and finance company (the founder of the DriveTime used-car finance empire). He noticed something: in the physical used-car auction lot — any car that was listed online with exceptionally detailed high-resolution photos — buyers would bid without ever seeing the car in person. So he posed a question: "What if you could buy a used car entirely online — bypass the physical dealership completely — and pick it up from an automated car-vending tower? Could that work?" And then Carvana was born — the world's first fully online used-car trading platform and the owner of the "car vending machine." Today Carvana is worth roughly 40 billion dollars.

Not automobiles. Not finance. Not artificial intelligence. Not autonomous driving.
It was "buy a used car online — and then it descends before your eyes from an all-glass automated car-vending tower — like a giant buying a kid a secondhand Lego car."

Ernie Garcia III was born in Phoenix in the 1980s. His father, Ernie Garcia II, was the founder of DriveTime — one of the largest integrated used-car, subprime lending, and auto-finance retailers in America — having built a multi-billion-dollar used-car retail and lending empire in a market of low-credit-score car buyers that banks and franchised dealerships had discriminated against. Ernie III grew up amid the exhaust fumes and plastic-chair smell of used-car auctions — and learned the detailed operations of used-car auctions and financing from the ground up — but he kept thinking: why do you have to stand on asphalt strewn with broken glass and oil stains, holding up a paddle to bid? The VIN (Vehicle Identification Number) and the data could handle the whole process online.

He launched Carvana in 2012 — carved out from segments of DriveTime's legacy inventory software and data — allowing customers to browse 360-degree high-definition images and Carfax reports for every vehicle online — complete the entire purchase online — and 3 to 7 days later, either have the car delivered to their doorstep or go to a nearby Carvana "car vending machine" and use a coin-sized token to trigger a car descending from the rails of a transparent steel tower.

Carvana went public on the Nasdaq in 2017. Its car vending towers (the tallest stands eight stories high and holds roughly 30 cars) became a social-media viral phenomenon — every customer who filmed themselves inserting a token and retrieving their car gave Carvana tens of millions of free impressions. During the pandemic of 2021-2022, used-car supply and demand both surged — Carvana's stock price rocketed — then in 2023, with interest rates rising sharply and integration difficulties after acquiring the Adesa auction network, the stock crashed over 90% — at one point facing bankruptcy. Then, after short-term debt restructuring, operational cost-cutting, and inventory reduction — it staged a dramatic rebound back to profitability.

Three things.

Thing one: Used-car auctions — can be completed on a phone screen — no physical lot required.
Carvana's online inspection and sales platform — is itself the elimination of "physical rent" for used-car auction lots. The per-vehicle turnover cost online is far lower than at a physical auction site.
In your industry — what things still "require a person to be there and see the physical object" — when in fact high-definition imaging and data trust could replace that?

Thing two: The car vending tower is not the primary sales channel — it is a billboard.
Most Carvana cars are not picked up from the tower — they are delivered directly to the home. But that vertical transparent car tower — it is nothing more than a physical trademark designed to make social media spread itself.
Would you consider building, near your company, a "physical installation that looks like a completely insane proposal" — just to tell the world that what you do is not what anyone else does?

Thing three: Don't be afraid of a 90% drop — as long as cash flow can turn back around.
Carvana's stock dropped from a peak of $370 to $3.50 in 2023 — then broke back above $200 in 2024 — a rebound of nearly 70 times. Used-car retail is cyclical, heavily dependent on lending and interest rates — but it is not going away.

Ernie Garcia III remains CEO to this day. His office is in suburban Phoenix — directly facing the old DriveTime headquarters building. Once, when a Morgan Stanley analyst asked him if he had ever thought about going back to practicing law — he smiled and said: "My father told me once — 'A good lot boy who stands behind the auction rail for three days understands auto finance better than a law student does in an entire semester.'"

← Previous Contents Next →
Chapter 195 United-Rentals

1997, Greenwich, Connecticut. Bradley Jacobs — a young man who had worked in M&A at Waste Management — decided to leave the garbage-hauling business and dive into a field that barely even counted as a formal industry: equipment rental. The American market for construction, industrial, and emergency-repair equipment — at that time was made up of several thousand small mom-and-pop rental shops — with no national chain brand whatsoever. Jacobs started with one small acquisition — and then accelerated, buying up every independent backhoe, lift, and generator rental shop he could find. Today United Rentals is the largest equipment rental company in the world, worth roughly 60 billion dollars.

Not construction. Not infrastructure. Not manufacturing. Not supply chain.
It was "a construction foreman — needs an excavator on the job site tomorrow morning at 7:00 — but he does not want to buy one."

Bradley Jacobs was born in 1956 in Connecticut. During his years in M&A at Waste Management, he witnessed firsthand how "continuously and relentlessly buying small routes" had turned a local garbage hauler — through hundreds of acquisitions — into the world's largest waste management company. He asked a very simple question: in the "equipment rental" market, which is twice the size of the waste-hauling market — why was there no national chain? The answer: every town's crane and excavator rental shop was run by a retired contractor who figured making enough to get by was fine — no ambition for national expansion.

Jacobs founded United Rentals in 1997 — raised capital from financial investors — and then began acquiring local equipment rental shops at a pace of dozens per year — rebranding all of them, unifying ERP and safety standards, centralizing equipment purchasing and parts procurement. His strategy was structurally identical to Huizenga's in the waste industry — "unify the back end — retain the local front-end manager — use a national inventory system to share surplus equipment across regions."

In the 2000s, United Rentals gradually evolved from "pure consolidator" to "full-lifecycle equipment manager" — it buys new equipment, rents it out at the highest rates during the early depreciation years, and then, before the equipment depreciates to salvage value, sells it into the used-equipment market — using the resale proceeds to offset the cost of buying new equipment — extracting profit from both rental income and residual-value recovery.

Then it layered on safety training and on-site project management services — for example, coordinating equipment allocation schedules across multiple trades on large project sites — effectively becoming a "mobile job-site equipment dispatch center." This is especially valuable during refinery turnaround cycles (shutdown maintenance) and post-disaster recovery — when all contractors and equipment must converge on-site within a very tight window.

Three things.

Thing one: Take an industry full of mom-and-pop shops — and brand it through relentless, continuous acquisition.
Jacobs took the acquisition playbook from waste hauling — replicated it in equipment rental — unifying brand, safety, procurement, equipment maintenance, and systems.
Does your industry have thousands of small independent operators in the same business — waiting for you to consolidate them under one brand and a unified management system?

Thing two: Rent out the equipment — but don't wait until it's scrap — sell it while residual value still exists, at the inflection point of rental rate cycles.
United Rentals is a revolving door — the equipment it buys is rented out at high rates during its "rental age" — and sold while meaningful residual value remains — and the proceeds from selling the old equipment go straight into buying the next batch of new equipment.
Are your assets being held onto past their "peak rental profitability" period — instead of being sold and the proceeds reinvested into newer, better equipment?

Thing three: Post-disaster recovery equals the ability to mobilize equipment nationally.
After a hurricane or a flood — every piece of local equipment is rented out — United Rentals' national inventory system can rush equipment from other regions of America to the disaster zone — renting it out at extremely high emergency rates. This "cross-state mobilization" capability — is something no local mom-and-pop shop can possibly match.

Bradley Jacobs retired as CEO in 2019 — but remains Chairman. When asked, "Why go into equipment rental — don't you miss garbage collection?" — his answer was: "Garbage and excavators share one thing in common — no one wants to own them — but everyone needs them tomorrow."

← Previous Contents Next →
Chapter 196 Truist-Financial

2019, the American Southeast. SunTrust Banks (Atlanta, tracing its earliest roots to a commercial bank and trust company founded in Atlanta in 1888) and BB&T (Winston-Salem, North Carolina, tracing its earliest roots to a small rural bank founded in 1872 that made harvest loans solely to farmers) — merged to form Truist Financial. The name is an invented word combining "True" and the suffix "-ist" — intended to convey "truth plus integrity." The new company quickly became the seventh-largest bank holding company in the United States — worth roughly 55 billion dollars.

Not fintech. Not Wall Street. Not the merger itself.
It was the shared DNA of the two most conservative banks in the Southeast — both built on the principle of "the person lending you money is the same bank manager you have known for decades" — and after over 140 years of independence each, they chose to come together to face the age of the digital behemoth.

The early incarnation of SunTrust was the Trust Company of Georgia in Atlanta — which in 1891 became the underwriter for the Coca-Cola Company — establishing a relationship between two profoundly different institutions (a sugar-water company and a conservative trust bank) that continues to this day. All of Coca-Cola's early public offerings were underwritten by SunTrust's predecessor, and SunTrust at one point held an enormous block of closely held Coca-Cola shares — a relationship sustained for over a century.

BB&T's early incarnation, meanwhile, was shaped by the father of branch banking, Alonzo Craig (who led BB&T from 1908 to 1950) — pushing "small harvest loans to farmers" across the tobacco and cotton fields of North Carolina — traveling between villages and farms by mule and horse every week — and collecting interest alongside a Bible lesson. This extremely conservative, profoundly decentralized rural-bank form — through hundreds of small-town savings-and-loan acquisitions over the decades that followed — grew into one of the largest regional banks in the Southeast — yet all the while preserved its mythologized "community bank culture."

The logic of the 2019 merger — SunTrust and BB&T faced the unlimited IT budgets, nationwide branch networks, and AI-driven risk pricing of Bank of America and JPMorgan Chase. If the two continued to fight alone, neither could survive independently in the long run of the digital banking race. The merged Truist built a single core banking system (gradually closing overlapping physical branches) — while redirecting the cost savings into online banking and mobile lending. Its headquarters is in Winston-Salem, inside the original BB&T tower — still the tallest building in town.

Three things.

Thing one: The Coca-Cola underwriter and the credit knight of the tobacco farmers — after 140 years, found themselves facing the same invincible enemy: the digital giants.
The shared threat to century-old banks — is not each other — it is the new financial applications with no physical branches. They chose to stand together — pitting "traditional trust" against "digital threshold."
In your industry — are there two legacy brands with different but complementary histories — that should merge to resist disruption by tech platforms?

Thing two: Merger equals merging systems — closing duplicate branches — but keeping the small-town bank face that "knows everyone's name."
Truist cut a large number of redundant branches in cities — but kept the long-standing branch managers in small towns — because they know that "locals need to see the same face when they take out a loan."
Can you cut the back office — but keep the "face-to-face face" with the customer?

Thing three: Coin a brand-new word for the name — don't hyphenate the two old names.
SunTrust + BB&T became Truist. After the merger, neither old sign remained — a new word was forged, forced to take root in the culture by sheer will. Immensely high risk — because it meant discarding two sets of century-old brand recognition at the same time.

Truist's first two CEOs, Kelly King (the last CEO of BB&T) and Bill Rogers (the last CEO of SunTrust), both quoted in their merger speech the same line from BB&T founder Alonzo Craig's 1910 notebook: "A bank should never feel colder than its poorest depositor."

← Previous Contents Next →
Chapter 197 Paccar

1905, Seattle, Washington. William Pigott — a mechanic who repaired small steam locomotives along the logging and coal-mine railway lines of the Pacific Northwest — partnered with a timber merchant to found the Seattle Car Manufacturing Company, and began building its first steam locomotives and later its first "iron truck chassis." They soon renamed the company Pacific Car and Foundry — specializing in heavy-duty custom trucks and railway cars for the western logging and mining industries. In 1918, the company acquired Kenworth — a small Seattle truck-body builder famous for hand-welding the frame rails of ultra-heavy-duty trucks. Later, Pacific Car and Foundry gradually became Paccar — today the parent company of Peterbilt and Kenworth heavy trucks, and one of the world's premier commercial truck and engine manufacturers, worth roughly 70 billion dollars.

Not trucks. Not engines. Not diesel. Not transportation.
It was one word: "The frame rail." From 1905 to today, every Kenworth and Peterbilt chassis longitudinal rail under the Paccar umbrella is pressed and heat-treated in a dedicated die from a solid steel billet — not fabricated from welded plate. This process gives its truck chassis a service life exceeding a million miles without ever needing a frame-rail re-weld.

William Pigott was born in 1869 in Missouri. He opened a mechanical repair shop early on beside logging railway lines — where he noticed that the longitudinal rails on the self-propelled steam track-tractors used to haul giant logs in lumber camps had an extremely high fracture rate. Because he had previously repaired bridge structures for the railroads — he replaced cast-iron frame rails with forged steel — and then redesigned the cross-section of that steel rail from the conventional C-channel to a rectangular closed-section, increasing torsional strength. That fundamental shape is still visible today on the Kenworth W900 long-nose heavy truck.

Through the 1920s and 1930s, Paccar kept acquiring: Kenworth (1924 — the remaining shares formally purchased in 1945), Peterbilt (acquired in 1958), DAF (1996 — one of the largest commercial truck manufacturers in the Netherlands). Each acquisition was left with its own brand, factory independence, and design language intact — but the back end — global component procurement, engine emissions certification, and captive finance (Paccar Financial) — was fully unified.

Paccar is not the truck company that sells the most trucks in the world (Daimler and Volvo Trucks have larger total volumes) — but it commands the highest per-unit price and the highest profit margins. A fully specced flagship Peterbilt 389 long-nose costs significantly more than competitors' equivalents. It holds an almost mystical status of loyalty in the hearts of North American long-haul truckers — the equivalent of Harley-Davidson's position in the motorcycle world.

Three things.

Thing one: Don't compete on "most sold" — compete on "one frame rail running a million miles without breaking."
Kenworth and Peterbilt chassis frame rails are forged and heat-treated steel — under extreme conditions they can exceed three million miles without fracture. Paccar has locked its heavy-truck brand premium into one single physical piece of hardware: the chassis.
Can your product withstand a physical test at the limit — three times the lifespan of competitors?

Thing two: A heavy truck is not a "vehicle" — it is the "driver's second home."
The sleeper cab on a Peterbilt long-nose — is where most long-haul drivers sleep every night. Paccar knows its buyer does not live in a house — they live in the sleeper berth — so the interior is designed to RV standards.
Is your product something the user "lives inside 24 hours a day" — rather than something they merely pass through and "use"?

Thing three: Keep Kenworth and Peterbilt as two fully independent design brands.
Paccar does not unify them — even though the internal chassis-part sharing rate is invisibly high — but the driver does not see the parts. What the driver sees: round headlights and a matte-black grille is a Pete — squared-off brass grille is a KW.

William Pigott passed away in 1934. In his lifetime he never drove a Kenworth or a Peterbilt — he only drove steam track-vehicles. But on the circular drive outside Paccar's Bellevue headquarters — there sits a gleaming black Peterbilt 389 and a deep-green Kenworth W900 — and beside the front axle of both trucks stands a brass plaque: "1905 — One Frame Rail."

← Previous Contents Next →
Chapter 198 Aflac

1955, Columbus, Georgia. John B. Amos — a local insurance broker who had spent years selling life and health policies — together with his two younger brothers, Paul S. Amos and Bill Amos, hung a new sign on the door of a small office in Columbus: American Family Life Insurance Company. Their first product was called "cancer insurance" — a specialized supplemental policy that paid a cash benefit only when the insured was diagnosed with cancer. The entire insurance world ridiculed the idea: "Why wouldn't you sell comprehensive health insurance — instead of just one disease?" The Amos brothers said: "Because cancer is the one disease everyone fears the most — and the money from comprehensive health insurance never reaches our policyholders." The company was later shortened to Aflac — and today Aflac is the most recognized supplemental health insurance brand in both the United States and Japan, worth roughly 50 billion dollars.

Not health insurance. Not life insurance. Not mutual insurance. Not government healthcare.
It was "insure only one disease — cancer — but the moment you are diagnosed, you get a check in your hands, and you can use it to pay your mortgage, buy food, do whatever you want — no need to explain to anyone."

John Amos was born in 1924 in the small town of Auburn, Georgia — his father was a local judge. He served in the Army Air Corps during World War II, earned his JD from Yale Law School, and returned to Georgia to work as an insurance broker. In the 1950s, he watched his clients go through every kind of misery with ordinary medical and comprehensive health insurance claims — but the one thing that sent every working-class family into total panic was: "The doctor said — it's cancer." The waiting periods, deductibles, and policy clauses that kicked in before a major illness meant they simply could not get their hands on disposable cash to cover the most basic non-medical survival needs — rent, feeding young children, transportation.

The first cancer insurance policy he and his brothers launched carried an extremely low annual premium — just a few dozen dollars — and if the insured was diagnosed with cancer — they received a cash check. There were zero restrictions on how that cash was used — it ran parallel to medical insurance reimbursement — no conflict, no impact on underwriting waiting periods — but triggered only by "the one disease everyone dreads most."

This concept was later extended by Aflac into accident insurance, short-term disability insurance, dental and vision insurance, and other supplemental lines — every one of them following the same design: "small premium, pays a lump sum or daily cash amount upon triggering event, runs parallel to primary medical insurance."

Aflac entered the Japanese market in 1974 — launching a nearly identical cancer supplemental insurance product — signing group contracts with Japanese corporate labor unions (especially at large manufacturing and trading companies) and deducting tiny premiums directly from payroll. Japan ultimately became Aflac's single largest market — contributing the majority of its revenue and profits — and cancer insurance became a default employee-benefit option at nearly every major Japanese company.

Then Aflac turned a duck into the most successful insurance mascot in the world. In 2000, comedian Gilbert Gottfried, in his piercing, grating voice, yelled "AFLAC!!!" — and a white duck would swoop in from nowhere to rescue hapless customers — this commercial ran continuously for two decades — making the AFLAC duck a trademark with the same universal name-recognition level as the Nike Swoosh.

Three things.

Thing one: Insure only the one thing people fear the most — not everything.
No one was offering cancer insurance in 1955 — because "insuring just one disease" sounded too narrow. But that one disease was enough to bankrupt a family. Amos was not afraid to do just one — because he knew it was "every person's deepest fear."
For your product — is "doing everything" perhaps not as powerful as doing "the one thing the whole population fears most"?

Thing two: Give cash directly to the patient — don't fight with health insurance.
Parallel cash — it does not argue with the medical insurance you already have. Paid separately, paid simultaneously. This "stackable but not a replacement" form of supplemental insurance — became a massive industry in both the United States and Japan.
Can you build a product that "does not fight the existing big system — but provides a parallel, direct cash stream"?

Thing three: The duck — using absolute absurdity to crush every boring green-shield ad from every financial company.
Aflac's duck did not explain the policy terms — it only screeched the "company name." But the name was remembered.

John Amos passed away in 1990. His son Daniel Amos took over as CEO that same year, ignited Aflac's explosion in Japan, and executed the "duck" brand revolution. Daniel installed a duck-call sound in the elevator at the Columbus headquarters — every time the elevator reaches the ground floor, the door lets out an "AFLAC!!!" — visitors startle and laugh, and then never forget it for the rest of their lives.

← Previous Contents Next →
Chapter 199 Realty-Income

1969, Maryland. William E. Clark — a private real estate investor who owned a few small retail properties in Southern California — together with several partners, decided to form an exceedingly strange REIT: they would buy only one type of property — freestanding restaurant and retail single-tenant buildings — and sign ultra-long "triple net lease" (NNN) contracts with tenants — contracts that stipulated: the tenant pays all taxes, all insurance, all maintenance, all utilities — the landlord collects a net rent check each month — with virtually zero operating expenses. And then they would distribute 90% of the rent collected to shareholders in the form of monthly dividends. The company has since paid monthly dividends continuously, without a single interruption, for over 640 consecutive months. Today Realty Income is worth roughly 50 billion dollars.

Not commercial real estate. Not retail. Not location. Not development.
It was "one rent check arriving every month, lasting over 50 years without expiration."

William Clark was born in 1922. He owned several freestanding restaurant properties in California — including early locations of a Wendy's and a Taco Bell — and he noticed that quick-service chain tenants almost never broke a long-term lease — because the location, the build-out, and the equipment investment locked the operator firmly in place — even during economic downturns, fast-food restaurants kept paying rent. So he joined with several Southern California real estate partners to found Realty Income — focused entirely on building a pure passive net-rent REIT that would "only collect rent, with near-zero operating costs."

All of their properties are "single-tenant" — one building, one tenant — on a triple net lease: the tenant is responsible for all three categories of operating expense (taxes, maintenance, insurance plus common area). Realty Income simply receives the net rent check the tenant mails in each month. The tenants are typically: Walgreens/CVS pharmacies, Dollar General, FedEx shipping centers, 7-Eleven, AMC theaters, Life Time fitness centers — "essential-service-type, e-commerce-resistant, high-renewal-rate" chain brands. These are usually standalone buildings with a parking lot — not inline storefronts inside a shopping mall (malls involve shared common areas and shared expense burdens — which carry operational and cost risks).

Realty Income's monthly dividend was branded in 1994 as "the monthly dividend company" — registered as a trademark — and it now pays a dividend to shareholders once a month, in the middle of every month. In its entire history — through recessions, COVID lockdowns, interest-rate swings — it has never once missed or reduced a payment.

Three things.

Thing one: Use the triple net lease — shift every operating cost onto the tenant.
Taxes (property tax) → tenant. Insurance → tenant. Maintenance → tenant. Utilities → tenant. Your return comes only from "the building itself" and the government-permitted physical address — you are utterly uninvolved in the operations of the restaurant or pharmacy.
Can you — through contractual packaging — shift all the "operational variables" in your business onto your counterparty — while you sit at the outermost layer collecting nothing but "pure rent"?

Thing two: Buy "Amazon-proof" properties — pharmacies, gas stations, fitness centers, convenience stores — things e-commerce cannot replace.
The e-commerce resistance of quick prescription pickup (CVS/Walgreens) and immediate hot food or drink (7-Eleven, McDonald's) — has been, so far, nearly absolute. Realty Income's tenant list is made up of "those physical necessities that Amazon cannot same-day deliver."
Are your assets resistant to "online substitution"? Can they achieve physical uniqueness?

Thing three: Monthly dividends — registered as a trademark — and paid continuously without interruption.
The monthly dividend is the ultimate promise in commercial real estate — Realty Income only holds properties that can generate extraordinarily high-certainty net cash flow, and is willing to give up property-appreciation speculation in exchange.

William Clark retired in the 1990s. Through the 2010s and beyond, Realty Income has continued to acquire properties by repeatedly issuing stock and debt — its portfolio has grown from a handful of fast-food locations to over ten thousand freestanding net-lease properties across the United States and the United Kingdom. Whichever small town you drive through in America — east, west, north, or south — when you pass a Walgreens pharmacy, the ground beneath your wheels very likely belongs to Realty Income.

← Previous Contents Next →
Chapter 200 WW-Grainger

1927, Chicago. William W. Grainger — a young manager working in electrical wholesale procurement in Chicago — noticed something peculiar. When his wholesale company sold a factory a generator (a big-ticket item), the customer would tack onto the same purchase order a casual inquiry: "Do you also have that special insulated copper busbar that goes between the generator and the distribution panel?" The big companies — GE, Westinghouse — only made equipment assemblies worth thousands or tens of thousands of dollars — they had zero interest in a small copper bar. But that small copper bar was exactly what the maintenance electrician needed most urgently, right then. So Grainger founded W.W. Grainger — "maintenance, repair, and operations" (MRO) industrial supplies distribution — which today has become a B2B supply-chain giant that stocks millions of industrial components, from screws to air compressors, in warehouses across the country and guarantees same-day delivery. Today W.W. Grainger is worth roughly 50 billion dollars.

Not manufacturing. Not IoT. Not supply chain.
It was "so that when your equipment breaks down, any single small part — you can find it in stock, today."

William Grainger was born in 1894 in Ohio. He started out working at an electric motor wholesale house in Chicago — selling large generators, motors, and transformers. He noticed that factory customers who bought a generator would ask at the bottom of the same PO: "Do you have that insulated copper strip that goes from the junction box to the busbar?" His sales colleagues always ignored that fine-print line — because the commission on that tiny item was not worth the time. But William saw the other side: if that small copper bar was out of stock, the installation of the entire generator would be delayed for weeks. So he left the motor wholesaling business and in 1927 founded W.W. Grainger — dedicating himself entirely to "the small parts no one else could be bothered to sell" — things like relays, fuses, push-button switches, screws, connectors, and belts.

His first Grainger catalog was a thin mimeographed list of motor repair parts, hand-typed in 1927. By the 1940s this handbook had grown into a thousand-page "industrial bible" — a copy sat on the desk of every factory electrician and maintenance foreman. Even after the internet age, a vast volume of industrial purchasing habits still begins with checking the Grainger website — its SKU count and inventory locations have grown to a scale that is nearly impossible for any new entrant to replicate.

Grainger's core competitive strength is maintaining an extraordinarily high in-stock rate — across its hundreds of local branch warehouses and large distribution centers — so that a maintenance technician who orders an essential part at 10 a.m. (say, a motor start capacitor with a specific overheat-protector rating) can receive it on the job site by 2 p.m. that same day. Industrial maintenance cannot wait — because every hour a production line is down costs tens or hundreds of thousands of dollars.

Three things.

Thing one: The small things that big companies don't want to sell — that is the monopoly business.
The generator giants (GE, Westinghouse) don't sell that copper busbar — but every generator they sell naturally creates a derivative demand for it. Grainger caught that derivative — and built an empire on it.
Do the strongest players in your industry — generate a massive volume of "small needs" every day that they cannot be bothered to fulfill?

Thing two: Catalog + local warehouse = the most trusted tool of the maintenance technician.
Grainger's first-generation product was its "book" — the product catalog. Now its product is the fulfillment network that guarantees "same-day or next-day delivery." The core has always been "minimize the time equipment stays dead."
Can your service — turn the customer's most expensive "downtime" into your value metric?

Thing three: Don't contract-manufacture. Don't make anything. Just do "mass distribution of any small part."
Grainger does not produce a single thing. But it distributes millions of items made by others — and it has the capability to supply them instantly. It is the ultimate form of the "industrial general store" — for every factory that faces the daily panic of "where do I find that small broken part."

William Grainger passed away in 1969. His company, to this day, still operates by the first business principle he established in 1927: "Never let a customer's machine stay down because of a small part you could have stocked." That principle is now inscribed on the floor of the company headquarters lobby.

← Previous Contents