"Deep Dive: The 1994 Founding"
Spring to summer 1994 — a statistic, a framework, a cross-country drive, and a garage in Bellevue.
The statistic
In the spring of 1994, Bezos encountered a number that reorganized his thinking: web usage growing at 2,300% a year. He has always credited the source — John Quarterman’s Matrix News newsletter. The number’s precision didn’t matter; its implication did: whatever you built on the internet would be riding an exponential. The question became what to build.
The framework
Bezos’s decision tool was the regret minimization framework: imagine yourself at 80, look back, and minimize regret. He concluded that not participating in the internet — “this thing called the internet that I thought was going to be a big deal” — would be the regret, while failing at a startup would not. The framework is worth noting because it recurs in his management writing: decide for the story you’ll want to tell, not the odds you’ll compute.
The list
He made a list of roughly twenty products that could be sold online — software, music, videos, books — and evaluated each on the economics of remote selling. Books won: a fragmented market with no dominant mail-order player, millions of titles (a physical superstore held ~175,000), and a standardized product where selection beat location. The famous gloss: the world’s biggest bookstore could only be built without a store.
The drive
In the summer of 1994, Bezos left D.E. Shaw, and he and Mackenzie drove from New York to Seattle — Mackenzie driving while he wrote the business plan on a laptop in the passenger seat. Seattle won for three reasons: the tech talent pool (Microsoft was there), proximity to the big book distributor Ingram (in Roseburg, Oregon), and — the small, telling reason — Washington State’s relatively small population, which meant fewer customers would trigger sales-tax collection obligations in the mail-order era.
The incorporation
5 July 1994: Cadabra, Inc. is incorporated in Washington State. Within months it became Amazon.com, Inc. — named for the Amazon River, the world’s largest, and for the letter A. (Relentless.com was also registered; it still redirects to amazon.com.) The garage was in Bellevue, in a rented house; early hires worked at door-desks, and a bell announced each sale until sales made the bell absurd.
The money
The first serious capital was family: Jacklyn and Miguel Bezos invested roughly $300,000 — much of their savings — in 1995. Bezos’s pitch to his parents, as he tells it, gave them a 70% chance of losing everything; they invested in their son. Seattle’s early tech network followed — including Tom Alberg and others — and in 1996 Kleiner Perkins Caufield & Byers, via John Doerr, invested about $8M.
The launch
16 July 1995: amazon.com goes live. The first book sold was Douglas Hofstadter’s Fluid Concepts and Creative Analogies — a fitting inaugural title for a company built on analogical thinking. Within weeks, orders were coming from all 50 states and 45 countries. The site’s pitch was selection: a million titles searchable, versus a superstore’s shelves.
Source notes
- 2,300% statistic and Quarterman/Matrix News attribution: Bezos’s repeated accounts (interviews, Academy of Achievement).
- Regret minimization framework and the age-80 framing: Bezos interviews; see key-ideas.md.
- The twenty-product list, books decision (175,000 vs. millions), Seattle rationale (Microsoft talent, Ingram proximity, sales-tax nexus): The Everything Store (Stone, 2013).
- The drive (Mackenzie driving, business plan en route), Bellevue garage, door-desks, the bell: The Everything Store.
- Incorporation (5 July 1994, Cadabra → Amazon, river rationale), Relentless.com redirect: The Everything Store.
- Parental $300,000 (1995; “invested in their son”), Alberg, KPCB/Doerr ~$8M (1996): The Everything Store.
- Launch (16 July 1995), first book (Hofstadter, Fluid Concepts and Creative Analogies): The Everything Store.
Limitations
- Founding narratives lean on Bezos’s and early employees’ retellings; figures like the $300,000 and the twenty-product list are consistently reported but originate with the participants.