(FORTUNE Magazine) – On the morning of Sept. 28, Amazon.com announced that on the morning of Sept. 29 it planned to make an "announcement significantly affecting the world of e-commerce." From Wall Street to Silicon Valley, the effect was immediate.

"I was desperate--calling around all day trying to find out what was going on," said Mark Rowen, an Internet analyst at Prudential Securities. At the Grand Hyatt in Manhattan, Forrester Research was hosting an e-commerce summit. Toby Lenk of eToys was scheduled to speak. So was Priceline's Jay Walker. Yet all anyone could think about was Amazon. "Here I am at our conference, and all we can do is talk about what Amazon is going to announce tomorrow," said Carrie Ardito, a Forrester e-commerce analyst. "Do you know anything?" Rumor had it that Amazon was opening a real-world store; it was going into business with a big-box retailer; it was launching clothes/travel/office supplies online. "I heard they were doing something with software," said Buy.com CEO Greg Hawkins. By the end of the day, on no news other than the announcement of the announcement, the stock market had boosted Amazon's value by $1.5 billion.

At nine the next morning in New York, a crowd of reporters and analysts crammed into the Versailles room of the Sheraton Hotel. The place was buzzing. "Do you have any idea what this is about?" Rowen pressed a reporter from TheStreet.com. She didn't, and asked what he knew. "I heard something about a partnership with Home Depot," he began, but before he could continue, Amazon CEO Jeff Bezos stepped onto the riser at the front of the room. The din stopped, and five TV cameras trained on him. "Sixteen months ago Amazon.com was a place where you could find books," Bezos began, hands folded behind his back as he paced the stage. "Tomorrow Amazon.com will be a place where you can find anything." With that, he introduced the latest installment of the Amazon potboiler: the serialized story of one company's ambitious plan to take over the world--the e-commerce world, that is.

Throughout the year, Amazon has been on the move. On average it has announced a major initiative every six weeks. In February it bought 46% of Drugstore.com. In March it launched online auctions--two days after rival eBay announced a secondary stock offering. In May the company took a 35% piece of HomeGrocer.com. In June, 54% of Pets.com. In July, 49% of Gear.com. That same month Amazon opened two new online shops: toys and electronics. Last month's announcement was Z-shops (an online mall) and All Product Search (a product browser). "We're always talking about what they're going to do next--they're constantly keeping us guessing," gushes Forrester's Ardito.

Forget about Amazon as the Wal-Mart of the Web. Bezos is aiming for something even bigger. So big, in fact, that it hasn't been invented yet. "I get asked a lot, Are you trying to be the Wal-Mart of the Web?" says Bezos. "The truth is, we're not trying to be the Anything of the Web. We're genetically pioneers. Everybody here wants to do something completely new. I wake up every morning trying to make sure I can confound journalists and pundits who try to encapsulate us in an eight-second sound bite."

In Bezos' vision, Amazon.com will be the center of the e-commerce universe. Books, pet food, tennis shoes, banjos--whatever e-shoppers want, they can buy it, or locate it, on Amazon.com. Picture Amazon as an octopus, its tentacles reaching out all over the Web. The potential payoff is huge. Investors certainly think so. After Amazon announced Z-shops and All Product Search, its stock rose 23%, to $80 a share. "This is so big, so important, that you have to be invested in it," says Morris Mark, a portfolio manager who added to his Amazon stake after the announcement. Pausing, he adds, "Of course, they've got to make it work."

That is precisely why the world (the e-commerce world, at least) is addicted to the Amazon saga. Can it pull everything together? Can it fuse all those great ideas into one great business? And can it ever make any money? "They have a great opportunity, but they could screw it up," says Henry Blodget, an Internet analyst at Merrill Lynch who made a name for himself last year predicting that Amazon's stock would hit $400. It seemed crazy at the time, but on a split-adjusted basis, he was right on target.

Amazon's four-year rise from upstart online bookseller to one of the largest retailers on the Web is now legend. The company has defined e-commerce as we know it. One-click shopping, customer reviews, online gift wrapping: Amazon invented it all. Go anywhere else on the Web--toysrus.com, homedepot.com, macys.com--chances are you'll find something copied from Amazon. It was the first place most people ever bought anything over the Internet. Now it is the largest online seller of books, music, and videos. It has 12 million customers and projected sales this year of $1.4 billion. Since going public in 1997, its stock has risen from $1.50 a share to $80. Its market cap is $28 billion--more than twice as big as that of Sears, five times bigger than Kmart's, 17 times bigger than Barnes & Noble's.

But as Amazon grows, so do its troubles. This year, while sales are expected to be up 136%, operating expenses will leap 252%, to $869 million (that includes acquisition costs). To pay for its expansion, Amazon issued $1.25 billion in convertible bonds earlier this year. Until those are converted to equity, the company's yearly interest expense is $93 million. It also now spends more than $300 million a year on sales and marketing. Meanwhile, its sales growth is slipping. Second-quarter sales grew just 7%, vs. 33% the year before. In 1998, Amazon's book sales--about 70% of its overall business--grew 275%; this year that number will drop to 82%. At the same time, revenues per customer are falling, from about $47 in 1997 to $29 this year.

All in all, net losses in 1999 will top $600 million. Even the most optimistic estimates say the company won't earn a penny for at least two years. And Amazon--the superstar stock--is beginning to look a little less glittery. Until Bezos' September announcement, it was trading at $60, 43% off its 12-month high of $105. "The losses have got to start coming down," says portfolio manager Morris Mark of Mark Associates.

Two years ago, Amazon had cyberspace largely to itself. Now the Internet is teeming with e-tailers. Buy.com--which sells books, videos, and music--is programmed to scan Amazon.com's prices and automatically undercut them. Yahoo's online mall rivals Amazon's, with 7,000 stores and over four million items, including books, clothes, music, and toys. Excite@Home bought iMall in July. And industry pundits have been predicting all summer that Wal-Mart's new site will steamroll Amazon. We'll have to wait until next year for that showdown: The new wal-mart.com won't launch before 2000.

On a recent sunny morning in Seattle, high on a hill overlooking Puget Sound, Amazon's competitors and mounting losses are the last things on Jeff Bezos' mind. He's trying to crank open one of the old-fashioned windows at his new headquarters, the former Pacific Medical building. "This is great--I love being able to open the window," he bursts out cheerfully. A boat horn sounds in the distance and he laughs--an infectious, gulp-from-the-throat laugh that he is known for. It's almost impossible to be in the same room with Bezos and not have a good time. He's relaxed, he's funny, and he's disarmingly humble. "Nobody is more surprised than me by what has happened over the last four years," says the man who in that span has amassed $9 billion.

Windows wide open, Bezos bounds across the room to a computer in the corner to demonstrate All Product Search. "Look, crampons!" he exclaims, pointing to a little square in the upper-left-hand corner of the site that reads SEARCH OF THE DAY: CRAMPONS. "That's something we don't sell. It's an oddball thing. Now people can find them here." Type in whatever you're looking for, he tells me excitedly--crampons, ponyskin loafers, bongo drums--and Amazon will scour the Web to find sites where you can buy it. Of course, first it checks to see whether you can buy the product on Amazon.com. Chances are you can: Amazon now offers 19 million items--everything from a 53-inch Sony Projection TV to the Discovery Channel Human Origins Deluxe Dig Adventure Kit.

To fully understand the retailing empire Bezos is trying to build, I go to Fernley, Nev. There's not much here except desert, rattlesnakes, and a massive Amazon.com warehouse--the first of seven it will open around the country by the end of this year. All in all, the company will add more than three million square feet of warehouse space at a cost of about $200 million. It is the first distribution network designed specifically for online retailing--designed, in other words, to send out merchandise item by item to individual customers. (Traditional warehouses send out merchandise in bulk to stores.) "There is nothing else like it," says Fernley's warehouse manager, Mike Perkins, as he leads me up a mesh steel staircase to an area high above the warehouse floor called Pick Delight. From here we survey the guts of Amazon.com's operation.

This is where it all happens, where hundreds of people--yes, real people, earning $7.50 an hour--make that Tom Clancy book you clicked on this morning appear on your doorstep tomorrow. It's also probably the most crucial element of Bezos' new plans. Delivering millions of items of merchandise is no easy job. To learn how to do it, Bezos went to the master. He hired Wal-Mart's former information chief Richard Dalzell. (Wal-Mart, renowned for its distribution system, tried to stop him. It sued Amazon for stealing trade secrets. But Bezos prevailed. The suit was settled last spring after Amazon agreed to some conditions, including the reassignment of some of Dalzell's underlings.)

The warehouse at Fernley is vast--more than 580,000 square feet. In the middle are rows of bookshelves crammed with bestsellers: Harry Potter and the Prisoner of Azkaban, Fat Busters!, Sue Grafton's O Is for Outlaw. Around the sides, giant steel-frame shelves hold stacks of DVD players, TV sets, VCRs. Above them is a stretch of shelves filled with Teletubbies, Furbys, and Barbies. As orders come in, workers known as pickers race around pulling the items off the shelves and loading them onto a giant conveyor belt that runs throughout the warehouse.

The belt eventually carries all this stuff to Pick Delight, a long aisle of empty cubbyholes. When you click the "buy it now" button, you are assigned one of the cubbies. The conveyor belt automatically ejects the items you ordered as it passes your cubby. As I watch, an Ozzy Osbourne CD slides into one cubby; a Panasonic camcorder goes into another; Mr. Potato Head shoots into one farther down. A few seconds later a PalmPilot III slides in next to Mr. Potato Head, and a blue light comes on at that cubbyhole. At this signal, a young woman clad in overalls rushes over, pulls out the merchandise, and puts it in a large cardboard box, which she then sets back onto the conveyor. The box disappears down the line, headed to the station where it will be packed and shipped. At full capacity, Amazon's new warehouses will be able to ship $10 billion a year in merchandise. As far as Bezos is concerned, Amazon is only getting started. "The goal is universal selection: 'earth's biggest river, earth's biggest selection,'" he says, chanting the company tag line.

Still, even he realizes that Amazon can't sell everything. That's where Z-shops and All Product Search come in. Z-shops is essentially an online mall; anyone--from your grandmother to a megaconglomerate--can set up a store on Amazon. The monks of Spencer Abbey, a Trappist monastery in Spencer, Mass., now sell their homemade tea and jam ($21.95 a package) on Amazon.com. Office Max uses the site to sell staplers, manila folders, file organizers, and tape dispensers. Each store pays a less-than-nominal $9.99 monthly fee, plus 1% to 5% commissions, in return for access to Amazon's 12 million customers. And Amazon gets to say it has 500,000 products available. "The idea is to let people find anything they might want to buy online," Bezos tells me. "[Amazon] is a 'Katrina Store' or a 'Jeff Store.' The notion is that you take customers and put them at the center of their own universe." It's a great idea, a huge idea. Can he pull it off?

Entering Amazon's new headquarters feels like walking into a giant pep rally. Photo collages celebrating each new Amazon launch hang in the entrance. In the reception area, trophies Amazon has won in such contests as Best Innovation and Top [Web] Site, line the window sills. Everywhere employees want to tell you what a good time they're having. Says David Schappell, a recent Wharton graduate with a 1950s-style buzz cut: "Everyone here is just charged up. We're passionate about what we do."

The instigator of Amazon's high spirits is Bezos. He has an incredible knack for winning people over. Last June, Joe Galli, the former president of Black & Decker's Power Tools division, had just accepted a top job at PepsiCo's Frito-Lay. When Bezos called to ask him to be Amazon's president, Galli wasn't interested, though he did agree to have lunch. "We had an instant chemistry," recalls Galli. "Sparks were flying. I had a ten-hour conversation with Jeff." By the end of it, ditching Frito-Lay was a no-brainer. "Bezos is such a visionary. [He's] changing the world, making history," says Galli. "I believe I have the most fabulous mentor on earth. To sit at the table with Jeff Bezos and learn about the Internet is incredible."

Having Bezos rally the troops is so crucial to the company that recently it started Radio Amazon, an internal station that plays tapes of Bezos talking about his plans. "Jeff can't be around as much anymore," says Joy Covey, Amazon's chief strategist. Radio Amazon, she says, is a way of keeping 4,000-plus employees plugged into his vision.

The trouble is, once you're outside Bezos' gravitational pull, that vision is hard to hold on to. Because the reality is that Amazon.com is a long way from succeeding with its big plans. It is struggling in every one of its new categories. In auctions, it's getting clobbered by competitors: eBay has three million listings; Yahoo, nearly one million; Amazon, at last count, just 140,000. In electronics, launched in July, it's off to a bad start: Pioneer and Sony, two of the biggest manufacturers, say they won't allow Amazon to sell their products. "Our main customers are our brick-and-mortar dealers," says Pioneer spokeswoman Joni Saphir. "Most of the manufacturers are trying to protect them." Indeed, both Pioneer and Sony have agreed to supply Circuit City's online store. When Amazon managed to get Pioneer's products through third-party dealers, Pioneer said it would take action against those dealers. Sony has announced a policy in which only authorized retailers may sell its products on the Web.

Amazon's toy shop has similar troubles. Some manufacturers, like Brio, maker of the popular wooden train sets, have refused to supply Amazon. Others, like Lego, are in "test mode" with the company, eager to see how well Amazon moves their merchandise at Christmas. "We just don't know if Amazon is the right place for us to be," says a Lego insider. Here, too, Amazon is going up against an entrenched online competitor. With a year's head start, eToys dominates toys on the Web and is not likely to give ground without a fight.

Even with its latest ventures, Z-shops and All Product Search, Amazon is in a tough spot. With All Product Search, it is, in effect, sending customers to the competition. With Z-shops, it risks losing control over its famed customer service. Seventy percent of Amazon's revenues come from repeat business because the company is "religious about the customer experience," says Bezos, who in a 45-minute interview uses the word "customer" 53 times. But now some guy with the online name "ferd1" in Milwaukee is responsible for sending you that Elvis clock you ordered from Amazon.com. If he (or any other Z-shopkeeper) doesn't deliver, or sends you crummy merchandise, Amazon's reputation will surely suffer. Of course, the company is trying to solve those problems, many of which it says are the growing pains of a new business. It offers customer reviews of each Z-shop merchant. In toys and electronics, it has hired managers from traditional retailers to help smooth relations with manufacturers.

For now, Amazon can afford to make mistakes. It has $1 billion in cash from its bond offering. So, even if the stock market were to turn on it, Amazon would have enough in its war chest to keep its plans on track. Still, just because Amazon can sell books does not mean it can sell everything else. More and more the question comes up: When will all this pay off? Bezos insists that the company can make money anytime it wants. "We were profitable in 1995, six months after we started," he says. (This was before Amazon's IPO, so there is no public record of his claim.) But, he continues, "that was the biggest mistake we ever made." He believes that concentrating on profitability now would mean missing out on "the big opportunities of the Internet." Although Bezos won't comment directly on Amazon's earnings' timetable, it will probably be years before it earns a dime. "Look at USA Today; it took 11 years to become profitable," he argues.

Bezos and his team say they've been upfront with investors from the start about profitability. "When it comes to Amazon, people fall into two camps," says chief strategist Covey. "Camp 1 is 'Ohmygod, they're not profitable.' Camp 2 understands what we're trying to build." In other words: Nonbelievers, invest your money elsewhere. And believers, well, trust us. "I recognize I'm going on faith here," confesses Amazon bull Henry Blodget.

How will the Amazon story end? Will it win the race? Conquer the world? Bezos himself isn't sure. "I don't want to give the impression that our future success is assured. I believe the opposite. I believe our future success is not assured," he says, adding with a grin, "If you look at the history of pioneers, it's not good." He even seems fascinated by his own cliffhanger. Everywhere he goes, Bezos carries a small digital camera. "I try to take at least one picture a day so that I'll have some hope 50 years from now of reconstructing my life," he explains, snapping my picture after our interview.

Right now, there seem to be two possible conclusions to the Amazon story. Ending No. 1 goes like this: In ten years Amazon becomes so huge, so omnipresent, that it will be hard to imagine that it started out as a tiny bookseller way back in 1995. Ending No. 2, equally believable: Amazon is undone by its own ambitions and winds up as a footnote in the history of business. In the meantime, Amazon watchers await the next big move. Sure enough, as this story goes to press, a news release appears in my e-mail: "Amazon's new shopping tool will be unveiled in November."