Amazon: How a Garage Bookstore Built an Empire

Part of Retailer Playbooks — history-first profiles of every company on the NRF Top 100 Retailers list.

Amazon lands at #2 on the NRF Top 100 Retailers 2026, the National Retail Federation's annual ranking compiled with Kantar, with $293.85 billion in 2025 U.S. retail sales. The number is almost too large to picture. The company behind it started as a single website selling books out of a Bellevue, Washington garage, run by a man who quit a Wall Street quant job after reading a statistic about how fast the web was growing.

Why Books, Specifically

Jeff Bezos left D.E. Shaw in 1994 with a list of twenty possible products to sell online, narrowed to five: books, music, videos, computer hardware, and software. He picked books for a specific, almost boring reason. There were more book titles in print worldwide than any physical superstore could ever stock, and the format didn't spoil, didn't need fitting, and didn't vary by size or color. A warehouse-free catalog business could out-select Barnes & Noble without out-spending it, according to Wikipedia's history of the company and FundingUniverse's company history.

He drove from New York to Seattle, writing the business plan on the way, and picked Seattle for its technical labor pool and its proximity to a major book distributor in Roseburg, Oregon. The company was briefly called Cadabra, after "abracadabra," before Bezos renamed it Amazon in November 1994 for the river, the biggest thing he could think of. The site opened to the public on July 16, 1995, offering a searchable database of more than a million titles from an inventory that, per FundingUniverse's history, held only about 2,000 books on hand. Most orders were filled through wholesalers, a workaround that let a company with almost no capital behave like it had a warehouse the size of the industry.

Losing Money on Purpose

Amazon went public in May 1997, an IPO of three million shares that reached one million customers and $147.8 million in sales by that October, and Bezos used the moment to say the quiet part out loud: the company would prioritize market share and long-term position over near-term profit. It was not a hedge. Amazon lost money for years by design, plowing IPO proceeds and later debt into distribution centers, opening an East Coast facility in New Castle, Delaware, in 1997 and expanding the Seattle warehouse by 70 percent the same year, chasing same-day-ship rates on in-stock orders.

That growth-over-profit stance had a body count. When the dot-com crash hit in 2000 and 2001, plenty of internet retailers with the same growth-at-all-costs playbook went to zero. Amazon laid off 1,300 employees and closed distribution facilities to survive, then posted its first-ever quarterly profit in Q4 2001, five million dollars on the books after seven years of losses. It was a company that had spent its adolescence building fulfillment infrastructure nobody was sure it would live long enough to use.

From Bookstore to Everything Store, One Category at a Time

The expansion beyond books followed a deliberate cadence rather than a single leap. Music and video arrived in 1998. Toys, electronics, and tools followed in 1999. The company folded in international footholds early, acquiring Bookpages in the UK and Telebook in Germany in 1998, and it built customer-facing features that were unusual for the time: shopper reviews, purchase-based recommendations, and an early alert service for new titles in a customer's interests.

The turning points that mattered most weren't retail moves at all. In 2002, Amazon launched Amazon Web Services, opening up the computing infrastructure it had built to run its own store to any outside developer who wanted to rent it. In 2006, Fulfillment by Amazon let third-party sellers store inventory in Amazon's warehouses and ship under Amazon's promise. Both were plumbing before they were products.

The Insight the About Page Won't Give You

Here's the part that doesn't show up in Amazon's own telling: AWS was not a strategic bet on cloud computing as a category. It grew out of Amazon's own engineering teams spending, by internal accounts, roughly 70 percent of their time on undifferentiated infrastructure work instead of customer-facing features, according to Wikipedia's account of AWS. Leadership's fix was to build one shared platform for the whole company and let engineers stop reinventing storage and databases for every new retail feature. Amazon then sold that internal fix to the outside world. By the first quarter of 2016, the same source notes, AWS had become more profitable than Amazon's entire North American retail business, contributing 56 percent of corporate profit that year on a fraction of total revenue. That inversion, tooling built to serve the store outearning the store, is the real Amazon story, and it's a pattern few of its retail peers have managed to repeat at anywhere near the same scale.

Prime, launched February 2, 2005, at $79 a year for free two-day shipping, did something similar for the customer side of the business. It didn't invent loyalty programs, but it monetized shipping cost as a subscription instead of eating it as a promotion, which meant the more a member ordered, the more Amazon made from a fee already collected upfront. By 2016 that had scaled to roughly 54 million U.S. members, per Wikipedia's entry on the program, and by 2021 to over 200 million worldwide.

What Survived the Climb

Bezos stepped down as CEO on July 5, 2021, twenty-seven years to the day after founding the company, handing the role to longtime AWS chief Andy Jassy. The acquisitions along the way, Zappos, Kindle-era hardware investments through the late 2000s, Whole Foods in 2017, MGM in 2022, read less like a shopping spree and more like a company continuously buying its way into whatever the next layer of infrastructure turned out to be: shoes and customer service culture, then groceries and physical real estate, then a content library for the video service bundled into Prime.

The garage bookstore's real invention wasn't selection. It was treating every constraint, warehouse space, shipping cost, server capacity, as a business to be built and then sold back to everyone else.

Retail's biggest stories are rarely about the products on the shelf. They're about the infrastructure quietly built to get products there at all, this series exists to tell those stories one company at a time.