For a time, no one seemed to mind very much. Certainly, Mr. Bezos appeared not to be worried by Amazon's losses. "There were no sacred cows with Jeff," said Venky Harinarayan, who was a top strategist at Amazon for two years, starting in 1998. "There was no conventional wisdom."
There also didn't seem to be an organizational chart, or anything that much resembled a personnel department in Amazon's first five years, said Jeremy Eskenazi, who joined the company in 1999 as its top recruiter but soon found himself running its human resources department.
"If Jeff is passionate about something, he pays attention to it," said Mr. Eskenazi, who left Amazon in 2001. "But H.R. and legal and accounting and all that more mundane stuff Jeff tended to look on as nonsensical, and he didn't support those areas."
Instead, Mr. Bezos focused on expansion. He directed an acquisition binge that he now concedes led to excesses that turned out to be as excruciating as "root canal without anesthesia." Among the more painful ones were his decisions to buy a majority stake in Pets.com in 1999, and then to pay $60 million for a large portion of the online delivery service Kozmo.com.
Both turned out to be legendary dot-com bombs that went out of business shortly after Amazon invested in them. Pets.com seemed perhaps the most foolish; among other things, pet food is very heavy, and ruinously expensive to ship.
"We believed that if we did not participate in these categories at that time, we would be forgoing the opportunity to participate in those categories forever," Mr. Bezos said. "We believed, incorrectly as it turns out - and this was my bad - that these were land-rush opportunities." Amazon took $350 million in losses between 2000 and 2002 for its failed investments in dot-com companies.
That same land-rush logic fueled Amazon's decision to start selling toys, electronics, tools, patio furniture and even cellphone service plans - all in 1999 and 2000. It also went head to head with eBay in online auctions, but it has yet to gain much traction after six years. More recently, Amazon has started selling jewelry, health products and musical instruments, bringing to 31 the number of categories on its Web site.
"I do think Amazon may have expanded too far, too fast," said Peter S. Fader, a marketing professor at the Wharton School of the University of Pennsylvania. Another Wharton marketing professor, Jerry Wind, added: "With Amazon, there is always a danger that it is spread too thin."
Aram Rubinson, a financial analyst at Bank of America, is inclined to agree. In a research report he published last November, he said Amazon should cut its inventory and sell only profitable items, even at the expense of revenue growth. "Other retailers, like Best Buy, Home Depot and Staples, have opted to slow growth in order to improve profits," wrote Mr. Rubinson, who has a sell rating on Amazon.
As is, books, music and videos - the first three categories Amazon entered - still account for nearly three-quarters of its revenue, and Ms. Johnson at Forrester said she doubted that Amazon would ever "make much of a dent" in many of the other markets it has entered.
Mr. Bezos scoffed at skeptics. Amazon has gold-plated customer service, he said, and has championed any number of nifty early innovations, like one-click buying. With such superior service, he seems to suggest, why won't the world beat a path to his door to buy everything from toasters to trumpets and tennis rackets?
"It takes a long time to teach 49 million active customers about our company and all we offer," he said. "Given it takes a long time, why not get into those areas as early as we can?"
Ms. Johnson offered one reason: Amazon has trouble focusing. "Frankly, there aren't the resources dedicated to each category," she said, "because Amazon in the meantime is launching more new categories, and getting into new services, and trying to develop a search engine and doing all these things trying to become all things to all people."
Read furthermore -> A Retail Revolution Turns 10 (6) Finish.