Jeff Bezos founded Amazon, the company that sells nearly everything and, less visibly, rents out most of the internet's computing power. He was born in Albuquerque in 1964 to a seventeen-year-old mother and a father who left; the man whose name he carries is Mike Bezos, a Cuban who arrived in the United States alone at sixteen and adopted him at four. He spent his childhood summers on his grandfather's ranch in south Texas, castrating cattle and fixing windmills.
He studied electrical engineering and computer science at Princeton, then went to Wall Street, ending up as a young vice-president at D. E. Shaw, a quantitative hedge fund that hired physicists to find patterns in markets. In 1994 he read that web usage was growing 2,300% a year, and decided that a number like that only happens once. He quit a job most people would consider the destination, drove west with his wife while writing the business plan in the passenger seat, and started selling books out of a garage in a Seattle suburb. His parents put in about $245,000, most of their savings, and he told them to their faces there was a 70% chance they would lose it.
The first order shipped in July 1995. The company went public in 1997 at $18 a share. Then the dot-com crash took the stock from over $100 to about $6, magazines ran the headline "Amazon.bomb", and analysts spent 2001 arguing over whether it would survive the year. It did, on razor-thin retail margins and a refusal to price for profit. The part that made the fortune came later and from an unexpected direction: in 2006 the company started renting out its own internal computing infrastructure to other businesses. Amazon Web Services now earns most of the group's operating profit, and a large share of the websites you use run on it.
He stepped back from running the company in 2021, keeping the chairman's seat. What he does with the proceeds is the rocket company Blue Origin, The Washington Post, and a ten-thousand-year clock inside a mountain in Texas. In the 2019 divorce, a quarter of his Amazon stake went to MacKenzie Scott, who has been giving it away faster than he has.
He took roughly 41% of a company he then deliberately ran at near-zero profit for two decades, financing its growth with customers' cash and suppliers' payment terms rather than his own money, and never sold, so the entire fortune is an unrealised mark on shares he still holds.
- Started with
- A professional household, adoptive father an Exxon engineer, grandfather a regional director of the US Atomic Energy Commission. Princeton, then senior vice-president at a Wall Street quant fund by his late twenties. He was walking away from a large guaranteed bonus, not from poverty. His parents put $245,573 into Amazon; siblings added $10,000 each.
- The decisive move
- Not founding Amazon, running it at structurally suppressed margin and refusing to convert ownership into income. Every dollar of gross profit went back into fulfilment, price cuts and infrastructure. Two consequences did the work: he never needed to sell, so compounding was never interrupted; and the reinvestment bought the distribution layer itself, turning a retailer into the toll booth other sellers pay.
- How the money was realised
- There was no event. His salary has been about $81,840 since 1998 and he has never taken a stock grant. The stake fell from ~41% to ~8.8% through dilution and divorce while its value rose from roughly $180m to a quarter of a trillion dollars.
- Elapsed time
- Thirty-four months to paper-rich at the 1997 IPO, and that is the misleading number. The stock fell about 94% between 1999 and 2001, the company carried a $2.86bn accumulated deficit, and the first full year of profit came in 2003. Call it nine to ten years to anything durable.
Transfers
- Finance growth with other people's working capital before your own. Customer pays at checkout, supplier is paid in sixty days. This is the single most copyable mechanic in the book, and it is more available in markets with informal cash and long supplier terms than in the West. Deposits, prepayment and subscriptions are all the same trick.
- Separate your income from your ownership. If your living costs are funded by the asset you are compounding, you will eventually liquidate it.
- Sell the tooling you already built for yourself. AWS is the famous case, but the mechanism is small, you solve an internal problem and the fact that you use it daily is what makes it good enough to sell.
- Own the layer everyone must pass through, not the product on top. Retail margin was the bait; the marketplace fee and the ad slot are the business.
Does not transfer
- Losing a quarter of a million dollars of family money and having the family be fine. Then $8m from Kleiner Perkins within twenty-four months.
- Never needing the money. "Don't sell" is a strategy available only to someone whose bills are already paid. A founder who must convert assets to eat is not making a worse decision, he is playing a different game.
- The window. Internet retail from 1994 with no incumbent and a public market willing to fund billions in losses on a narrative. Running unprofitable for a decade today gets you shut down, not crowned.
- Surviving 2001. That was not skill; it was having enough cash and enough lender patience to outlast a crash that killed hundreds of equally committed founders.
That the lesson is "be willing to lose money for a long time." It is not. Amazon's retail business was cash-generative at the working-capital level almost from the start, because customers paid before suppliers did. The reported losses were a deliberate accounting outcome of pushing that cash into assets. He was not burning investor money hoping to be rescued, he was converting one form of cash into another. Copy the suppressed margin without the negative cash cycle and you simply run out of money.