One Click: Jeff Bezos and the Rise of Amazon.com Part 4

Most observers-some within Amazon-thought the idea was crazy. Allow compet.i.tors to sell through Amazon and compete with products Amazon was also trying to sell? But Bezos knew that people would be able to find other products through the Internet, albeit less easily than finding them on Amazon. By letting them sell through Amazon, he ensured that Amazon remained the best way to find products. And once again conventional wisdom was wrong and Bezos's wisdom was right. Forrester estimates that in the last quarter of 2010, Marketplace accounted for 35 percent of the company's revenues for that quarter.

Not all of Bezos's new retailing ideas worked. In March 1999 he added Amazon Auctions to compete with eBay, and lost. It turned out that eBay was too tough a compet.i.tor. But that doesn't detract from his position as a great entrepreneur. The Silicon Valley mantra is to experiment, keep the things that work and drop those that do not. The key is to know when to hold and when to fold.

Also in 1998, Bezos started expanding beyond the United States. His first international sites were in the United Kingdom and Germany. With the Internet taking off, Bezos had decided he once again had to move quickly, especially now that he had money sitting in his public stock just waiting to be tapped. He explained the moves at the time with a sense of urgency, revealing that it was always his intention to expand beyond books:Our product extension and geographic expansion is better late than early. Why better late than early? We had to first focus on the book business and grow that until we were comfortable with it. There are always numerous opportunities to expand. We try to err on the side of being slow. Fortunately, we are not capital constrained, but we are definitely people constrained. We only pursue opportunities when the people bandwidth is not constrained.... The single most important criterion that we use to acquire a new company is this: Who are the people behind this venture, and what is the people bandwidth of the acquired company going to be? We are looking for business athletes indoctrinated in this s.p.a.ce and companies that have a culture that is common with ours.

When Bezos started moving overseas, German book publis.h.i.+ng giant Bertelsmann AG thought a partners.h.i.+p would make sense. Bertelsmann CEO Thomas Middelhoff made an offer. He even sent a corporate jet to fetch Bezos from Turkey, where he was vacationing. Middelhoff's driver went to the airport to pick up the American billionaire when he arrived in Germany, and was surprised to find a casually dressed Bezos toting a bright yellow backpack.

Middelhoff and Bezos met for four hours. Middelhof wanted to create a fifty-fifty joint venture with Amazon in Europe. That deal could have conceivably given Amazon better access to Bertelsmann books, perhaps at a discount, as well as books from other publis.h.i.+ng houses the German giant owned, including Random House. Bertelsmann also had a joint venture with America Online in Europe, which could have been another advertising venue for Amazon. Bezos turned the deal down. He didn't say why publicly, simply stating that they couldn't settle on a deal. Middelhoff told the press that he thought Bezos didn't want Bertelsmann to have too much influence. "Jeff was nervous about giving up control," he said.

So Bertelsmann instead offered $300 million for a 50 percent stake in Barnes & n.o.ble's Web site, Barnesandn.o.ble.com. "This venture has one purpose-to compete with Amazon in the U.S.," said Middelhoff.

That deal was like marrying the ugly stepsister, since in the previous six months Barnesandn.o.ble.com had sold just $22 million worth of books, while Amazon had sold nearly ten times that amount. But some observers thought-once again-that the deal could mean trouble for Amazon's expansion plans. Bertelsmann had deep pockets, and all the advantages that could have gone to Amazon would now go to Barnesandn.o.ble.com. Fortune magazine asked in a headline, "Is the party over for Amazon?"

Middelhoff put Jonathan Bulkeley, who set up AOL's British operation, in charge of the new venture. The New York Times said that Bulkeley's job was "to make Mr. Bezos as miserable as possible." Middelhoff was certain he would succeed. "The future will show who will be better, Jeff or Jonathan," he said. "Personally, I bet on Jonathan."

For his part, Bulkeley talked tough. He noted that Bertelsmann had 4.5 million t.i.tles to offer, compared with Amazon's database of 3 million. Barnes & n.o.ble also tried to buy book distributor Ingram, Amazon's biggest supplier of books. "We're motivated. We've got a killer team," noted Bulkeley.

Bulkeley resigned in January 2000, thirteen months after taking on the job. He said he wanted to spend more time with his family and his own investments. The resignation came just after he reported that sales of Barnesandn.o.ble.com had tripled in the previous quarter, to $81.5 million. Amazon's fourth-quarter revenues had increased more than three and a half times, to $250 million. It's not impossible to compete with Amazon, it's just hard to catch its momentum. That fact can be credited to Bezos's determination to get to market first, with the best service, and stay ahead of the compet.i.tion.

Bezos kept up his mad pace of expansion through 1999, buying or investing in a new company or entering a new business almost every month. He invested in Drugstore.com in February. In March he started the auction business, bought e-Niche (a site for buying and selling music), MusicFind to expand CD sales, Bibliofind for rare books, retailer Accept.com, a company called Alexa Internet, which tracked what sites people visited and recommended others they might find interesting, then added zBubbles to tap Alexa's database in order recommend where to buy products (including those from Amazon) whenever people visited a site reviewing products. That month he also tried to buy Blue Mountain Arts, an online greeting card business, but was rebuffed, so he started his own compet.i.tor to Blue Mountain. In May he invested in HomeGrocer.com. In June he invested $45 million in 250-year-old auction house Sothe-by's to add high-end items to his auction business. Direct sales of consumer electronics, toys, and games started in July, the same month he bought almost half of sporting goods retailer Gear.com. In September he invested in Della & James, which ran an online wedding gift registry site. He started selling home improvement products, software, video games, and jewelry and leather goods (through a $10 million investment in Ashford.com). He added a Gift Ideas feature in November and signed an agreement with Next-Card, Inc., in order to offer an Amazon-branded Visa card. That was also the month he announced zShops, the precursor of Amazon Marketplace, to let businesses and individuals sell products through Amazon.

With that last announcement, he also changed Amazon's logo. Instead of a logo in which the letter A in Amazon was designed to look like a flowing river, it became simply the word Amazon, with a swooping arrow underneath to point from the letter A to the letter Z in the name. The idea was that consumers could now use Amazon to buy any product, from A to Z, even if Amazon didn't sell the product itself. And if it did sell a product but someone else had it cheaper, Bezos would let customers buy it from the lower-priced retailer through Amazon.

He explained it at the time as an almost altruistic feature for customers. "In the categories where we are selling things directly, if we can't be compet.i.tive then we shouldn't be standing in the way of our customers," he said at the zShops announcement. However, he also made it clear that he wouldn't stand in the way of his customers as long as they didn't try to leave the Amazon store. "We really don't care whether we sell something through zShops, or sell something directly ourselves; it is sort of a wash for us," he added. "You can't sell everything on your own. You need to band together with third parties."

But the big message was this: No matter what anyone wanted to buy, they could probably get the best price by going through Amazon. If it wasn't the absolute best price, the difference was probably worth the security one got with an Amazon transaction. It's a reputation that Amazon still holds today.

There was, however, one bit of negative backlash to Amazon's incredible dominance. With the purchase of Alexa and Amazon's ability to recommend products that it knew its customers might like, people began to realize that the online retailing giant was collecting an astounding amount of information about their buying habits and interests. That made some people feel just a bit queasy. In January 2000, a San Francisco man filed a lawsuit complaining that Alexa was sending his personal information to Amazon without his permission. At the same time, Internet security consultant Richard M. Smith filed a complaint with the Federal Trade Commission saying that Amazon collected more personal info than it let on.

Amazon had to start disclosing its practices more openly in privacy statements that n.o.body ever read in order to quell the rising controversy. In fact, Amazon is still one of the most impressive machines for collecting data about people who use its services. These days, however, Google has displaced Amazon as the anti-privacy bogeyman. (When Google announced its bookselling business in 2010, privacy advocates even complained that would give the company too much information about what books people are buying, seemingly having forgotten about Amazon's database altogether.) Still, throughout 1999, Amazon's stock price continued to soar like Icarus trying out new wings. Wall Street a.n.a.lysts and others who kept warning about the company's lack of profitability were dismissed as old economy pessimists who didn't understand the new economy of the Internet. In April 1998, Keith Benjamin, an Internet a.n.a.lyst with BancAmerica Robertson Stephens in San Francisco, enthusiastically called Amazon "the poster child of Internet commerce." Amazon was also the poster child for Henry Blodget, an Internet a.n.a.lyst with Merrill Lynch, who became famous for his seemingly uncanny projections of the amazing growth of Amazon and other Internet stocks.

But the number of skeptics was also growing. There were plenty of knowledgeable people who believed the Internet economy was a bubble that had to pop someday. The question was when it would happen. But when it did happen, they felt, Amazon would be vulnerable. In 1999, the company reported a loss of $720 million, and had ama.s.sed $2 billion in debt, which was costing the company $125 million a year in interest. How could Amazon continue without the ever-soaring stock price, which provided Bezos with cash and equity to build warehouses, add inventory, and buy more companies in his grandiose dream to create the greatest retailing firm on earth?

A commentary article in a March 1999 edition of The New York Times by book author Peter de Jonge summed up the skepticism: "For all its all-nighters and tattooed punks humping books in the distribution center and golden retrievers wandering the halls in the corporate office," De Jonge wrote, "Amazon.com is a $20 billion, 2,100-employee company built on the thin membrane of a bubble, and this brings a manic precariousness to the place that no amount of profitless growth can diminish." In November 1999, Barron's magazine named Amazon (along with Microsoft) the most overvalued stock on the market.

n.o.body really knew which view was right. But by the end of 1999, everyone was sure of one thing: Jeff Bezos had forever transformed the business of retailing. He claimed that Amazon now sold eighteen million different items through its site. In December 1999, Bezos was named Time magazine's "Person of the Year." Bezos had reached the pinnacle of an industry he had created, and it seemed like n.o.body could touch him. By that time, the mountain that he lorded over like Moses on the Mount had already started crumbling beneath his feet. The only milestone he hadn't accomplished yet was the one he claimed wasn't important: He still hadn't turned a profit.

Chapter 11.

The Crash.

In June 1999, Black & Decker executive Joseph Galli accepted an offer to become president and CEO of PepsiCo's Frito-Lay North America Division.Then Bezos got to him. In a long meeting, Bezos convinced him to kiss off Frito-Lay and join him atop Amazon's summit as president and chief operating officer. After the purchase of mail-order tool-seller Tool Crib of the North, Galli helped Amazon get into the business of selling tools online, a much higher-margin business than the others Amazon had entered recently. Contractors could order tools on Amazon and have them s.h.i.+pped directly to their job sites.

That was the fun part of the job. The following January, Bezos, at least, said he was still having a great time. "I'm just plain having fun at Amazon.com," he said. "I'm a change junkie, and I can't imagine an environment more changing than the Internet in general and Amazon.com in particular."

Well, the Internet was changing, all right. After the Y2K scare turned out to be less threatening than a two-year-old's Halloween costume, everybody stopped upgrading the computers they had previously feared would crash, sending airplanes plummeting from the sky. Instead, sales of computers and other tech products did the plummeting, starting January 1, 2000. Within a few months, as the terrible sales started showing up in quarterly financial reports, tech stocks followed suit, and the dot-com bubble imploded like stars into a black hole.

Amazon's crash coincided with the dot-com crash, which became apparent after Amazon reported its first quarter results in April 2000 (the first quarter after Y2K-induced buying stopped). The crash certainly contributed to the collapse of Amazon's stock price. But it does not account for the sudden slowing in Amazon's revenues. After all, most of Amazon's revenues came from non-tech products, so the end of the tech-buying binge should not have hurt Amazon as much as other tech companies.

But signs of trouble at Amazon began even before Y2K hit and ran. Even as late as December 1999, a.n.a.lysts were projecting a net loss for Amazon of as much as $350 million. But the loss came in at a whopping $720 million, even though revenues had more than doubled, to $1.6 billion.

The year 2000 merely exacerbated the problems that had started at Amazon in the year Y1.999K. After years of annual reports to shareholders that began with a letter full of hyperactive hyperbole about the previous spectacular year, Bezos began his letter for the year 2000 annual report with: "Ouch. It's been a brutal year for many in the capital markets and certainly for Amazon.com shareholders." After years of triple-digit growth, revenues in 2000 grew just 68 percent, to $2.8 billion. Further, Amazon reported a $1.4 billion loss, compared to a $654 million loss the year before.

For almost any other multibillion-dollar company in the world, 68 percent growth would have been spectacular. But Amazon had grown too big in too short a time. Bezos's strategy of growing as fast as possible worked, putting the company into a position that made it almost impossible for anyone to catch (even today), at least in retailing. But the company had also started to spin out of control, plagued by inefficiency and overcapacity. With everything in high gear and the gas pedal to the floor, it was not able to stop quickly enough to avoid slamming into the wall that suddenly appeared around the Y2K bend.

At the end of 1999, for example, Amazon's most recent warehouse, an 850,000-square-foot behemoth in Coffeyville, Kansas, was only 10 percent utilized. Bezos was antic.i.p.ating that the company's continued torrid growth into new markets would fill the warehouse with new products soon enough. He was still pursuing the strategy that it was better to have too much capacity now than not enough when needed, which would delay the delivery of products and anger customers. As Bezos had once quipped, like the joke about obstetricians, "One baby dropped on its head is too many."

His ambition truly had no bounds, his head full of new ideas on what to add next. By mid-1999, he was talking about Amazon getting into travel services, banking, insurance, and other businesses far afield from retailing. And what if other Web sites came along to offer even lower prices than Amazon? "Members.h.i.+p clubs!" he said in December. "If you want to see all the information we collect on Amazon-the customer reviews, the professional reviews and use our agenting technology-you have to pay $30 a year."

But at the same time, increased compet.i.tion was also taking its toll, despite Amazon's rapid growth and domination of new markets. In fact, Bezos apparently knew that trouble lay ahead as early as the summer of 1999. Revenues were more than doubling, but losses were also growing in step with his ambitions.

That started worrying more Wall Street stock a.n.a.lysts, who had been blown away with Amazon's growth and stock trajectory until then. When Amazon held its quarterly conference call with a.n.a.lysts in June 1999, the Street was less impressed than usual. "Everyone thinks [the problems] all happened in April of this year [2000]," recalled Kelyn Brannon, Amazon's former CFO, in a December 2000 story in Fortune magazine. "We saw the turn-how the market was going-after the June [1999] conference call. The tone of the calls, the questions, the whisper conversations that went on afterward-it was a different tone. Rather than saying great quarter, great revenue growth-we stopped getting so many 'greats.' You get right into hard questions: Can you talk to me about direct margin? Can you talk to me about the operational efficiencies in your distribution center?" Brannon left Amazon six months after that conference call.

Galli was the one who had to straighten things out. In January 2000, he had to play the heavy and lay off 150 people in order to start putting Amazon on a path toward profitability. He hired new managers who knew how to run a company more efficiently, implemented tighter budgets, requiring all major purchases to be approved by top management. He became the symbol of Amazon's transformation from a very cool new-age Internet company to a more b.u.t.toned-down corporation that was actually p.r.o.ne to recession like every other company in the world. He even took away the employees' free aspirin and Tylenol-a move that created such an outcry that he had to restore it a week later. Employees were getting a lot of headaches.

As the recession deepened in 2000 and Amazon's stock plummeted like spit off a bridge, Galli himself began to need some of that Tylenol. In July 2000, a little more than a year after he had joined Amazon, the company announced that he was leaving. He took over as president and CEO of VerticalNet, an Internet company that ran dozens of industry-specific Web sites. Galli and Bezos insisted that things were fine between them, and that Galli decided to change jobs to get closer to his children, who lived with his former wife in Baltimore. VerticalNet was in Horsham, Pennsylvania. But VerticalNet was (and is) no Amazon. It never recovered from the Y2K fiasco, and in 2007 was sold to an Italian cement company.

Wall Street wondered if there was more to Galli's departure than family needs. For one thing, the previous CEO of VerticalNet said that he hadn't even been looking for a new CEO, that Galli had approached him about the job. And the timing of Galli's departure was odd: He quit the day before Amazon announced its second-quarter earnings. "Well, it's never a good sign when your president resigns a day before earnings," said Mark Rowen, a market a.n.a.lyst at Prudential Securities, the day Galli resigned. He suspected trouble. "The truth is, nothing goes from a small business to an extremely large business without any hiccups," Rowen said. "And I think we're seeing a hiccup at Amazon."

What Amazon suffered was more of a seizure than a hiccup. It's just that most Wall Street a.n.a.lysts were hoping Bezos would still pull Amazon out of its revenue spasm. But when he announced that Amazon had taken in just $578 million in sales for the quarter, $22 million less than some Wall Street a.n.a.lysts had hoped for, six a.n.a.lysts immediately downgraded the stock. Because of cost-cutting measures, Amazon's net loss was less than Wall Street had expected, but stockholders were hoping that increasing revenues would help it generate a profit that year. The stock dropped 13 percent that day. In total, thirteen a.n.a.lysts downgraded Amazon's rating in the year 2000. Between mid-December 1999 and the end of 2000, Amazon's stock lost 90 percent of its value, bottoming out at about $15 per share.

Was Amazon now toast? People had finally given up on unprofitable Internet companies, and Amazon was the biggest money-loser on the Internet. In the previous five years, Bezos had borrowed $2 billion-and had lost $1.74 billion of it. Wall Street started calling on Bezos to grow up and start running the company like a business, not a gambling casino. "It's time for these guys to start performing like real retailers," said Gene Alvarez, a retail a.n.a.lyst at research firm Meta Group. "The Internet panache has worn off and now it's time to start performing."

Bezos took the criticism to heart. He complained that he had gone from "Internet poster child to Internet whipping boy." At the same time, he finally changed the tune he had been singing-or perhaps whistling in the dark-for years. "Get Big Fast" was dead. He was now going to work on turning a profit.

To be fair, this s.h.i.+ft was, in a sense, all part of his plan. During a Q&A session with employees at a spanking-new distribution center in December 1999, he said he would switch to a strategy of working toward a profit when the "cone of opportunity" on the Internet had sufficiently narrowed to make it difficult for newcomers to squeeze through ahead of him, to gain a foothold in all the new markets he was contemplating. Bezos just didn't realize that the cone would suddenly close altogether.

Once that happened, though, Bezos adopted his new religion with all the fervor he had previously used to wors.h.i.+p the G.o.d of growth. A December 2000 article in Fortune magazine noted that he mentioned the word profit twenty-five times in an interview with the magazine. One of the mentions: "We have, for the first time, set an internal goal with the date for when the company as a whole is going to be profitable," he declared.

Bezos even showed the Fortune reporter an email he had circulated to the company's seven thousand employees to prove it. It read: "We're putting a stake in the ground: We're going to become profitable. That's right: We're aiming to have sales of $5 billion, produce over $1 billion in gross profits, and achieve solid operating profitability by . . ." (Bezos had blotted out the date, since he had a policy against making forward-looking statements. But the reporter got the date from another Amazon employee: Christmas 2001.) Bezos did what any stockholder-fearing CEO would do: He slashed costs by laying people off and cutting spending. His inspirational posters preaching about growth were replaced with something more down-to-earth: "Get the c.r.a.p [Can't Realize Any Profit] Out." He also started running Amazon more like a retailer than like a dot-com company, with both eyes on the bottom line. He closed unprofitable lines of business, wrote off bad investments, increased cost-cutting, and started demanding more realistic budgets. Every division head had to meet with him weekly to go over the division's budget. Each executive had to submit budgets with specific revenue goals, and timelines by which they would be met. Amazon managers started taking Finance 101 courses at the company's Seattle headquarters.

Bezos also tried to rea.s.sure Wall Street that he was cracking the cost-cutting whip at managers. At the third-quarter conference call with stock a.n.a.lysts, he said he had been telling managers that "we want you to find half a million dollars, we want you to find $750,000, we want you to find $1.5 million." Those rea.s.surances failed to inspire confidence. After the meeting, Amazon's stock price dropped 17 percent. People hadn't heard this kind of backtracking from the always gung-ho Bezos before.

That didn't mean, however, that he put the brakes on Amazon's expansion into new retail products. In 2000, he added sales of lawn and patio furniture, health and beauty aids, and kitchen products to the mix. He also made investments in living.com, Audible.com, and online car-sales company Greenlight.com. These companies, however, also paid Amazon hefty fees to be featured on Amazon's Web site, still prime virtual property on the Internet. Bezos's ambitions had not abated one bit. "The question now becomes, how much of the $5 trillion worldwide [retail] market is addressable?" Bezos said during the Wall Street conference call.

In October 2000, Amazon started showing improvement, this time giving Wall Street a pleasant surprise. Its quarterly loss came in at 25 cents a share. Most a.n.a.lysts had been expecting a loss of about 33 cents a share. There were other good signs: Operating losses were 11 percent of revenues, half what they had been a year earlier, while gross profit margins were up to 26 percent, compared to 20 percent a year earlier. Bezos had finally learned the trick most public companies had lived by for years: Never surprise Wall Street unless it's with good news. The stock regained about 30 percent of its value over the next three days.

It wasn't the end of Amazon's troubles, though. The company still struggled with a tough market throughout 2001, and its stock again receded. In early 2001, he laid off another thirteen hundred employees, about 15 percent of the workforce, including customer service workers, closing down a service center in Seattle. He also shuttered a warehouse in Georgia.

During this time he started another business that brought in immediate revenues: building and running Web sites for other companies, including Toys "R" Us, Target, Circuit City, and even book retailer Borders.

But those deals were done, in part, to make up for declines in Amazon's core business. In the third quarter of 2000, for example, Amazon sold $400 million worth of books, music, and videos. A year later that had dropped to $351 million. One a.n.a.lyst estimated that quarterly revenue per customer had dropped from $31 per employee to just $18 per customer in late 2001.

The Toys "R" Us deal, announced in August 2000, was a particularly big s.h.i.+ft in strategy for Amazon. The deal was set to last ten years, during which Amazon would help the toy retailer rebuild and run its site, Toysrus.com. The toy company hadn't done a particularly good job on its own, suffering through many site crashes and late deliveries. Amazon would handle the purchases, s.h.i.+pping, and customer service, but Toys "R" Us would own the inventory, which meant Bezos didn't have to spend money buying products up front. Toys "R" Us took the risk of paying for the inventory-including buying Amazon's toy inventory-while Amazon would stock all the toys sold online in its underused warehouses. That made things much easier for Amazon, since toys usually had to be ordered from Asian manufacturers six months in advance. Bezos didn't want to be in the position of having to write off more unsold inventory. This deal was the start of Amazon's business of selling products from other retailers through its site.

The deal, however, didn't last ten years. While it was hailed as a good deal for both companies at the time, getting them both out of the problems they were facing in online toy sales, both companies wanted out once their problems were behind them. In 2003, Amazon started allowing other toy retailers to sell through its site. In 2004, Toys "R" Us sued for $200 million in damages and the right to get out of the deal, saying that Amazon had agreed to sell only its toys through Amazon.com. Bezos countersued, a.s.serting that the toy retailer had failed to provide Amazon with enough inventory to meet demand. In 2007, a court ruled that Toys "R" Us could terminate the agreement, but awarded no damages. The previous year, electronics retailer Circuit City ended a similar relations.h.i.+p with Amazon after four years.

But in 2001, Bezos had his eye on a much closer horizon. At the fall earnings conference call, he said he expected to start making a profit by the end of the year. He was, however, very specific, promising only a pro forma net profit (a profit before most expenses are taken into account). He kept cutting back in the fourth quarter, reducing operating costs by half. That allowed him to further lower prices on books, music, and videos, with the hope that people would buy more.

It worked. On January 22, 2002, he made the conference call with a.n.a.lysts to reveal the results of the fourth quarter of 2001. Not only had he made the pro forma net, he reported the company's first net profit after all expenses were taken out. It wasn't much: just $5 million, or a penny a share. But it was much nicer than the $545 million net loss he'd reported a year earlier. This was to become a key contributor to Amazon's future profitability. By running the company more efficiently, he could keep prices as low as possible, cementing Amazon's reputation as the go-to place for online retail.

Try as he might, he could barely contain his enthusiasm. "As pleased as we are with this quarter, we have a ton of work to do," Bezos said during the conference call. Still, listeners could hear Amazon executives applauding and slapping high fives with each other at the end of the call.

They had reason to be happy. In four days, the stock rose 42 percent. That was still only a price of $14.44, but a huge gain for anyone who bought Amazon near the bottom.

And now that he had proved, at last, that he could run a profitable business, he continued Amazon's expansion, albeit at a less frenetic pace. In June 2002, he opened for business in Canada. In September he started selling office products. In November he started selling apparel. But his most important new business would not arrive until 2007. It was a new piece of hardware developed just for Amazon, called the Kindle.

Chapter 12.

Bezos Bets Big on the Kindle.

The CEO of the largest bookstore on earth believes that books have had "a great five hundred year run." But now, he insists, "it's time to change."

Some people may still love the musty smell of old books, the crispness of print on paper pages, or the comfort of a solid book like an old friend. Bezos disagrees. He's ready to toss them into the recycling bin to make way for new technology. "I'm grumpy when I'm forced to read a physical book because it's not as convenient," he complains. "Turning the pages . . . the book is always flopping itself shut at the wrong moment."

Bezos insists-and others agree-that the Kindle is simply a better form of book. Once again, he resorted to the strategy he first used when starting an online store: making the product something unique rather than simply imitating the physical version. "You can't ever outbook the book," Bezos said, "so you have to do things that you can't do with a book, such as in-stream dictionary lookup, changing fonts, and wireless delivery of content in 60 seconds. We have to build something better than a physical book."

That Bezos would so readily reject the paper product that started his company should not be a big surprise. He's an electronics guy, a fan of products written in base two computer code: 00 01 10 11 100 101 . . . It's the final content that's important, whether delivered with type on a page or with bits shot through cybers.p.a.ce to an electronic reader. And manufacturing and delivering electronic books is much, much cheaper than manufacturing and delivering the paper kind.

But it took another brilliant entrepreneur to give him the idea. Except for Web services, Bezos sold only physical goods in Amazon's first decade. Then, on March 4, 2003, Apple's Steve Jobs demonstrated that some physical products were unnecessary. The music CD was simply a way of delivering the real product, the music itself. But music can be digitized and s.h.i.+pped over the Internet without the cost of the physical CD or the expense of mailing. The iTunes music store was born.

Sometime in 2004 Bezos had an Amazon executive approach Gregg Zehr, a hardware developer who had worked at Apple and at palmOne, which created the Palm personal digital a.s.sistant. The executive asked Zehr to start a new company in order to create a new electronic book reader for Amazon. Zehr reportedly asked why he should be interested. The answer: "To change the world."

The start-up, called, mysteriously, Lab126, was kept very quiet for years. It is located in Cupertino, in the Silicon Valley, where Zehr recruited other developers from Apple and Palm. Lab126 would only reveal on its Web site that it was working on a "groundbreaking, highly integrated consumer product." One blogger who only identified himself as "j" came across the company's Web site in early 2006 and speculated that Lab126 was creating a compet.i.tor to Apple's iPod, or perhaps a smart phone. The mystery was solved on November 19, 2007, when Bezos announced the first Kindle e-book reader.

But is Bezos really that tired of physical books? One can never tell when his hyperbole is real or when he's just trying to sell a product. In October 2005, for example, even as Lab126 was getting started on the Kindle, he introduced a new feature that would allow customers to access pages of a book electronically through the Web site for a few cents a page. Or, for $1.99 per book, they could read as many pages of a book as they wished through their computer. At that time, he insisted that reading electronic books and paper books were separate products. "For the kind of book where you have long reading sessions, you will still want the physical book, but there may be times when you want to access [short sections of] a book for reference purposes," he said at the time. Or, he suggested, impatient customers waiting for the book they just ordered to arrive from Amazon may spend two bucks to get started immediately. One likely scenario is that he was testing the idea of paying for access to digital books while the Kindle was being tested.

Amazon was not the first company to offer electronic books. In fact, people had been toying with the idea for decades before the Kindle arrived. In 1968, Alan Kay, then at Xerox Corporation's Palo Alto Research Center, conceived of the Dynabook, a graphics-based portable computer (a concept which Steve Jobs later borrowed to create the Macintosh). But Kay also saw the device as an e-book, a way to download and carry around digitized books.

That meant people would have to have access to digitized books. So in 1971, Michael Hart, a student at the University of Illinois, Champaign-Urbana (where the Mosaic Web browser was later invented), started Project Gutenberg, the first project to digitize and archive written works, mostly books in the public domain. The project now has some thirty-five thousand works in its archive. The effort was an attempt "to encourage the creation and distribution of eBooks."

In the 1990s, companies started offering e-book devices in earnest. A company called Digital Book, Inc., sold floppy disks containing fifty digitized books in 1993. Within a few years, now-forgotten e-book readers such as Cybook, RocketBook, and SoftBook were reaching the market. Most sold in the range of $300 to $500, although one, the Everybook-which had twin facing color screens so that it could open like a real book-sold for $1,600. Online retailers started offering books in electronic form for sale in 1999. Even Microsoft got into the game by announcing a software program called Microsoft Reader, designed specifically to read electronic books, in 2000. They were all commercial failures.

Still, the field kept progressing. Random House and HarperCollins started digitizing some of their books in 2002. Google started digitizing public-domain books in 2004. Sony Corporation launched its first e-book in 2006. A year later, the Sony Reader was widely considered a dud, which Sony denied, although the company refused to reveal any sales numbers.

By 2006, people were beginning to question whether the public even wanted digital books, even as Apple's iPod popularized the business of downloading digital music through the Internet and directly into a portable device. "Consumers have proven time and again that they would prefer to buy and keep physical books," Evan Wilson, an a.n.a.lyst with Pacific Crest Securities who covers Sony, told BusinessWeek magazine in September 2007.

In early 2007, the rumors began to circulate about Amazon's coming e-book reader. Sony tried to turn up the heat in August by adding software from Adobe that would allow its Reader to download electronic books from other sources. It dropped the price by $50, to $250. It took out ads in publications such as The New York Times, USA Today, and Vanity Fair. It offered first-time buyers credits for one hundred free cla.s.sic t.i.tles. Those tactics didn't work either.

Then, on November 19, 2007, Bezos stood before a crowd of reporters at the W Union Square Hotel in Manhattan to announce the Kindle, an electronic book-reading device that he would sell for $399. Bezos has never said how many of the devices were available when the Kindle was first announced, but he did make a point of saying that they sold out in five and a half hours. It remained on back order for months. The online magazine TechCrunch cited a source claiming that by July 2008 Amazon had sold 240,000 of the devices. The book publis.h.i.+ng industry was irrevocably changed.

The early versions of e-books, it turned out, simply had too many flaws. Some people didn't like the design of the products. The Sony Reader was criticized as having a confusing interface with too many b.u.t.tons. In order to download books, it had to be connected to a computer. But the big problem was a severe lack of the most important aspect of an e-book reader: the electronic books themselves. Sony's product, for example, could only pull books from Sony's online service called Connect, which had just about one-tenth the number of t.i.tles available at a good-sized bookstore.

Bezos overcame those flaws with his usual approach: attention to detail. The Kindle is simple to use. It includes wi-fi, allowing it to connect directly to Amazon without going through a separate computer. Once an electronic book is purchased from Amazon, it downloads seamlessly within seconds into the Kindle (and, more recently, other devices the buyer designates, such as the iPad, iPod, iPhone, and Android phones). Simply click or tap on the t.i.tle of the book on the screen and the e-book opens to the last page you've read. It leverages the technology of electronics and the Internet to do things paper books cannot, such as allowing customers to tap on a word to call up its definition.

One of the important issues that Lab126 dealt with was the complaint that reading text on a computer screen for long periods tires the eyes. So Lab126 incorporated into the Kindle technologies called "electronic paper" and "electronic ink." First developed in the 1970s at Xerox Corporation's Palo Alto Research Center (PARC), it mimics the look of ink on paper by using electrically charged particles, black on one side and white on the other, which could be selectively flipped over with electrical charges to create black letters on a white screen. The Kindle-like Sony's earlier e-book-used technology from a company called E Ink Corporation. Founded in 1997 as a spin-off of the Media Lab at the Ma.s.sachusetts Inst.i.tute of Technology, it developed a slightly different approach using charged particles.

The e-ink display contains millions of "microcapsules," each one filled with white particles with a positive charge and black particles with a negative charge. The microcapsules are sandwiched between two layers of electrodes. Most of the screen keeps a negative charge on the top layer of electrodes and a positive charge on the bottom, pulling the white particles to the top of the microcapsules and the black to the bottom. That creates a white background on the screen. In order to form letters, the polarity of the electrodes is reversed in specific spots, pulling the black capsules to the top and forming black letters. Since the page is formed with particles, and not electrons, it looks like particles of ink on a page. And the particles don't dance like electrons being constantly refreshed. Once the page is formed, the particles remain in place without applying more power to it, saving battery life and creating stable, unblinking letters. All it needs is a quick flash of power to form the next page.

Sony had many of the same features, but not as well implemented. Its first readers apparently used e-ink technology that wasn't yet ready for the best-seller list. When a page was erased and a new one created, it produced an annoying flash between pages. The technology from Lab126 didn't have the same problem.

Bezos also had a big advantage: With access to virtually every publisher in the world, he could offer ninety thousand e-books when the Kindle was first released in 2007. The original device could store two hundred books at once.

Despite its technology and Bezos's claims of quick sales, the press was as skeptical of the Kindle's potential as they were of Amazon's ability to fend off compet.i.tion a decade earlier. The Kindle was ugly. (TechCrunch accurately described it as something that "looks like it came out of the 70's.") At $400, it was too expensive, as was the price of $10 per book. And it seemed as though it would be difficult to make the e-ink technology cheap enough to bring down the price significantly. "E Ink's technology isn't enough to make Kindle the breakthrough e-reading device that I and thousands of other e-book fans have been waiting for," wrote the respected technology journalist Wade Roush on the technology site Xconomy.

These days, Bezos is on a zealous campaign to popularize his Kindle e-book reader device and relegate the paper book to the dusty shelves of library archives. He has repeatedly said that the Kindle is the best-selling product on Amazon. But his definition of "product" is obscure. It's difficult to believe, for example, that Amazon sells more Kindles than it does the e-books to read on them, or even paper-based books. In terms of revenues, however, the Kindle may be a bigger product, since one Kindle costs $200 to $500, while the electronic books are usually $10 to $14 apiece.

But the Kindle has unquestionably had an impact. Since it was released the question among publishers has gone from "Do people really want electronic books?" to "Do people want to read physical books anymore?" Jeff Bezos single-handedly turned publis.h.i.+ng upside down with the Kindle. By December 2010, e-books were accounting for up to 10 percent of the revenues at some large publishers, despite the fact that the publishers sell e-books at half the price of hardcover books. That implies that just after three years on the market, up to 20 percent of books at some publishers are sold in electronic form.

Bezos is so avid about e-books, in fact, that he has been willing to sell them at a loss. In order to keep his cut-rate $9.99 price for most e-books, Bezos is reselling them at a loss of up to five dollars for each discounted book he sells. Bezos clearly sees it as the future of the business that got Amazon started, and is determined to remain the leader in this category by keeping compet.i.tors from getting a foothold on the virtual bookshelf.

But it's getting harder for him to maintain his pricing strategy. Bezos has tried to use muscle against publishers to force them to sell electronic books to him more cheaply, just as he has done with paper versions. Publishers fear that customers will become accustomed to cheap e-books (a fear that might already have some truth behind it), forcing them to put a wholesale price on e-books that will not allow the publishers to make a profit. So they are switching to a new pricing structure for e-books. In the new "agency model," publishers are demanding that they set the retail price of the e-books, generally $12.99 to $14.99, just as they do for the paper versions. The publishers will keep 70 percent of the price to themselves. The retailers can then offer discounts if they want, but they have to take it out of their 30 percent.

Bezos does not like this scenario and has been doing everything he can to stop it. On January 28, 2010, Macmillan CEO John Sargent traveled to Seattle to give Amazon his proposal for the agency model. If Amazon did not accept it, Sargent told the executives that they could keep their old model, but that would result in (as he wrote in his blog) "extensive and deep winnowing of t.i.tles" available to Amazon. Within a week, Amazon had removed all Macmillan books-paper and electronic-from its site, except those sold through Amazon by third parties.

The tactic failed. Bezos had to give in to Sargent's demands and within another week had restored Macmillan's books to the site. This was, no doubt, due partly to the fact that Steve Jobs had already agreed to the agency model, which could have given Apple better access to e-books from placated publishers. On the other hand, in October 2010 Amazon offered to pay royalties of 70 percent to authors who self-publish through the Kindle store, compared to 25 percent from most publishers.

For now, the Kindle still leads the market for e-book readers at its current price. Research company ChangeWave estimated that the Kindle had the largest share of the market in early 2011, at 47 percent. Apple's iPad (which does much more than just read books and is more expensive) had a 32 percent share. The Sony Reader and the Barnes & n.o.ble Nook were laggards, with just 5 percent and 4 percent of the market, respectively.

But the Kindle may not hold its lead forever. Before Apple's iPad was released, Amazon practically owned the e-book reader market. And other compet.i.tors keep trying. Barnes & n.o.ble has come out with a color version of the Nook, to compete with the black-and-white Kindle. At the end of 2010 Barnes & n.o.ble echoed earlier statements from Bezos when it announced that the Nook Color has become the best-selling item it has ever sold. Aside from its own stores, Barnes & n.o.ble also sells the Nook through other retailers such as Best Buy and Wal-Mart. The company said it sold more than a million e-books on Christmas day 2010.

And Google is making it tough for Bezos to strong-arm publishers for heavy discounts on e-books. In December 2010, it launched its own online store, called Google eBooks, to sell all those books it has been digitizing itself. Google is supplying applications to allow people to buy and download its e-books to every appropriate device, from the iPad to smart phones, except one: the Kindle. That could either limit Google's market (if the Kindle remains popular) or force Amazon to open the Kindle to e-books sold by other vendors (the format Kindle uses makes that difficult today).

Publishers, once resistant to Google's book-digitizing project, are now embracing Google's efforts, since Google has agreed to the agency model. Even brick-and-mortar stores love Google's approach, because they can now start selling e-books themselves. Google is allowing independent bookstores to sell its e-books through their own Web sites.

Retailers are confident, at least for now, that Google will not try to take the whole market for itself. "Google's business model is not to be a retailer," says Oren Teicher, CEO of the American Booksellers a.s.sociation. He believes Google will allow independent bookstores to more effectively compete against Amazon. "The fact that the cost of technology has come way down means that you don't need to be an international monolith to use technology," he says. The bookstores, be believes, will do a better job of recommending new books to their online customers, just as they do their store customers. "We're good at putting the right book in buyers' hands. We have knowledge and pa.s.sion about books."

Bezos, on the other hand, has a pa.s.sion about electronic business models. He's sure to keep advancing the Kindle in order to stay at the top of the market. His strategy has been to focus on a dedicated e-book reader, not a general-purpose device. That allows him to use technologies such as electronic ink, which refreshes too slowly for computers-it takes about as long for the next page to appear as it does to turn the page of a paper book-but is better for long-term reading and is still legible in bright sunlight. It's not certain, however, if that tactic will last. With new compet.i.tors racing into the market like an invading army, his early lead in the market might not sustain its momentum.

There's definitely a huge future market in e-books. Market researcher Forrester estimates that people bought $1 billion worth of e-books in 2010, and that by 2015 that will grow to $3 billion annually. The question is whether Amazon should focus on the Kindle or the e-books themselves. Rather than discounting the e-books, one argument says, Bezos may be considering eventually making the Kindle free. He has already dropped one version to $139-cheaper, as his TV ads say, than a good pair of designer sungla.s.ses. In October 2009, blogger John Walkenbach graphed the declining price of the Kindle, and noticed it was on a straight-line trajectory that pointed to zero in the second half of 2011. Author and blogger Kevin Kelly asked Bezos about that trend line in August 2010. Bezos smiled and said, "Oh, you noticed that." And then smiled again.

Then Michael Arrington at TechCrunch came up with a business model that would make it possible. In January 2010, Amazon made a great offer to select customers: Buy a Kindle, but if you don't like it, get a full refund-and keep the device. Arrington believes that it was a test run to see what the economic outcome of a free Kindle would turn out to be. He quotes "a reliable source" that says Bezos wants to give a free Kindle to every Amazon Prime subscriber, Amazon customers who pay $79 a year to get unlimited free two-day s.h.i.+pping and one-day s.h.i.+pping for $3.99 per item.

Amazon Prime subscribers are the company's best customers, those most likely to buy a lot of e-books from Amazon, covering the cost of giving away a free reader.

For now, Bezos isn't talking, but he clearly sees e-books as a huge part of Amazon's future. Market researchers estimate that Amazon sells three out of every four e-books sold. The question is whether the Kindle itself will be a part of that future. After all, Amazon also provides free applications that allow people to turn their computers and cell phones into e-readers in order to buy and download e-books from Amazon. People can even get free e-books-cla.s.sics no longer under copyright-from Amazon's store. Besides, although the Kindle is one of the most important products Bezos has introduced in years, it isn't the only one. Although he hasn't started offering insurance policies through Amazon.com (yet), he has been figuring out new ways to make money without selling physical products. His ambition still knows no bounds.