Ripping up the rules of management

Talk about giving away the store. In 2001, Amazon.com (AMZN) founder Jeff Bezos began offering free virtual shelf space to his competitors, letting them sell their wares alongside his own. Wall Street thought he had lost his mind. Even some of his top lieutenants thought it was crazy, especially since Amazon was pursuing what seemed like a reckless expansion plan -- selling not just books but everything from consumer electronics to discount pet supplies -- and still hadn't made a penny of profit.

"It was the kind of thing that became controversial internally," Bezos says. "It made people nervous. But the reality is that if you give customers what they want -- price, selection, and fast delivery -- you're going to get more sales."

Luckily for Bezos, that's exactly what happened. When he gave his competitors a piece of the action, his own business took off. Sales grew by 34 percent in 2003, reaching more than $5 billion. At the end of that year, after nearly a decade of operating at a loss, the company finally posted its first profit.

It was a classic Bezos move: Give something away for free, but do it in such a way that it ends up expanding the business. That's because he was one of the first people to stop thinking of the world of retailing as bricks and mortar, and to see it instead as databases and fulfillment mechanisms. You might think that having books shipped to you would cost more than walking down to the neighborhood bookstore, but Bezos realized early on that shipping costs were factored in no matter where you bought your bestsellers. (Not to mention the backlist titles that most bookstores couldn't afford to stock.)

Now Bezos is exploring yet more crazy ideas -- and making Wall Street nervous all over again. He's shifting his focus from the storefront to the back office, spending millions on Web services that look a lot like money-draining distractions.

Take Amazon's Simple Storage Service, or S3, which charges 15 cents per gigabyte per month to store data on the company's servers. It seems miles away from his core business of retail. But Web 2.0 startups have flocked to Amazon to save money and get to market more quickly.

Linden Labs, for example, has signed on with the service to store a Second Life software client. Even a behemoth like (MSFT) Microsoft finds it cost-effective to store its Direct Student Download program at Amazon's data centers. Bezos is also renting out computing power for 10 cents an hour through his Elastic Compute Cloud, or EC2. His litmus test: "If we needed this kind of technology internally, then chances are a bunch of companies could also benefit."

And having given away online shelf space to competitors, Bezos is making physical room for them in Amazon's warehouses. His new service, called Fulfillment by Amazon, lets smaller companies send their inventories to Amazon for distribution; now, when a customer places an order, it's Amazon that gets the automated message to ship the requested item.

Deutsche Bank Securities analyst Jeetil Patel thinks Bezos's Web services strategy is not as strange as it seems. "Amazon has always been a hybrid business, and this allows Amazon to take advantage of its existing technology," he says. "But it's unclear whether it'll be profitable or bring in significant revenue, or even how it'll evolve."

So far, the new ventures are bringing in only a trickle of cash. But that's fine with Bezos. "We take a long-term approach, and we're willing to invest for a number of years in new initiatives, which is one way we differ from other folks," he says. "We think of it as planting seeds, and we let those new trees grow."