Amazon: Bezos told his parents there was a 70% chance they would lose it all. The prospectus records it in two lines

In May 2001, Jeff Bezos said this in an interview: when he asked his parents to invest, he told them he thought there was a 70% chance they would lose all of it. He said it four years after Amazon’s listing, just after the dot-com bubble burst.

The same investment appears in the 1997 IPO prospectus. There are no odds and no conversation there. In February 1995, 582,528 shares went to his father. In July of the same year, 847,716 shares went to a trust whose beneficiary was his mother. Both were priced at $0.1717 a share. Multiply it out and you get about $100,000 and about $146,000, roughly $246,000 together.

The investment as told and the investment as recorded concern the same money and feel nothing alike. What follows lines up the prospectus, Bezos’s own letter and the interview, and rebuilds Amazon’s first two years from the numbers. Listed companies fall outside this site’s criteria for cases, so only the founding period is treated here, as a column.

From incorporation in July 1994 to 614 staff at the end of 1997

In the right-hand column, only the odds, the statistic and the story of the first orders come from the interview. Everything else is in the prospectus or the letter.

The 70% figure lives only in words spoken after success

The 70% odds appear nowhere in the prospectus. The prospectus does have a risk factors section, laying out continued losses and fierce competition at length, but that language was aimed at investors at the time of listing, not at what his parents heard in 1995. The only record of what he told them is his own recollection, four years after the listing.

There is no particular reason to doubt it. The role the number plays is clear, though. Told to an audience that already knows about the success, 70% becomes the weight in a story of “they invested anyway.” The paper records the same investment with just two numbers, shares and price, and says nothing about how ready the investors were to lose it.

Even so, setting the two side by side yields something. Against the $0.1717 a share his parents paid, the same prospectus lists Kleiner Perkins buying preferred stock at $14.05 a share and an offering price of $18. Per share, the venture capitalists paid about 82 times what his parents did, and IPO investors about 105 times. Before the listing, the investment he had said would be lost with 70% probability had become stock that professional investors paid more than 80 times as much for.

The founder lent the company money before anyone put money in

The lines that stood out most in the prospectus came before the parents’ investment. Three interest-free loans from Bezos himself are listed: $15,000 in July 1994, $29,000 in November and $40,000 in November 1995. That is $84,000 in total, down to the month of repayment. The first came back in August 1995, the second in April 1995, and the third was repaid the same month it was lent.

Founding stories cast the founder as someone pouring in their own money. On paper, the founder first lends the company money and recovers it within months of the parents’ investment arriving. What follows is inference: in choosing a loan rather than equity, there appears to be a design that separated his stake in the company from his own cash from the start. The desk thinks these three lines record the founder’s judgment better than the 70% story does.

First-year revenue was $511,000, with 11 employees

Revenue in 1995 was $511,000 and the net loss $303,000. There were 11 employees at year-end. The store opened in July, so these are half-year figures. In 1996 revenue grew about 31-fold to $15.746M, and the loss widened about 19-fold to $5.777M. In 1997 revenue reached $147.8M, about 9.4 times the previous year.

In his recollection, orders came from all 50 states and 45 countries within the first 30 days. The prospectus does not carry that figure, but half a year of revenue at $511,000 does not contradict it.

His own letter does not state the loss

The 1997 letter to shareholders is built from numbers Bezos chose himself: 838% revenue growth, 1.51 million customer accounts, repeat customers at over 58% of orders, 614 employees, 285,000 square feet of distribution space. It says the company will prioritize long-term value and take cash flows over how the accounts look.

Yet the year’s net loss appears nowhere in the letter. It acknowledges a business incurring losses but gives no amount. The prospectus had to state losses, and the letter did not. For the same company in the same period, changing the kind of document changes which numbers appear.

People who open their own books do it before any law requires it

Amazon’s founding-era numbers can be read in such detail because a listing brought statutory disclosure. A small business with no such obligation keeps its books private unless the owner opens them.

kan.kikuchi published the 5.91 million yen annual revenue of VR games he built alone, along with a development cost of 420,000 yen, 813 hours of work and a 14.1% refund rate. The maker of Streets of Rogue disclosed starting with $18,000 of his own money. Neither number was written because a law demanded it.

The list setting Amazon beside other legendary companies is in the column tracing 22 founding legends to their sources.

What could be confirmed

The limits need stating. What he told his parents exists only in his 2001 interview. The prospectus amounts are products of share counts and prices, and the totals themselves are not printed. When Kleiner Perkins bought its preferred stock could not be pinned down in what was read. Per-share prices are compared as printed in the prospectus, so stock splits follow the prospectus’s basis. The numbers in the 1997 letter were chosen by its author and lack a prospectus’s completeness.

Two lines of share allocation outlast the 70% story

Follow Amazon’s founding years through numbers and only the 70%, the heart of the story, sits in words spoken after success. Share counts, prices, the months of the loans and the first year’s loss lie flatly in the listing documents.

The legend repeats the former, and what can be verified is the latter. The latter is also usually what records what the founder actually decided. Lend, get repaid, sell equity to the family. Three lines of loans adding up to $84,000 describe this company’s first decisions most precisely.