(FORTUNE Magazine) – Deep in the bowels of eToys, a typical virtual retailer, the stuff that makes the business tick is about as unvirtual as you can imagine. One week before Christmas, the cold, cement-floored warehouse at this privately held e-commerce startup looks like a Toys "R" Us superstore without the bright lights and teeming crowds of shoppers. Instead, guys at PCs are printing out customer orders. So-called pickers, armed with clipboards and shopping carts, are rushing up and down the aisles, pulling merchandise from a stock of 10,000 different items. For each order, a picker picks. A checker checks. A packer packs. A wrapper wraps, if the customer has requested it. None of this looks very high tech. Seeing the Internet merchant from the inside, I feel like Dorothy unmasking the Wizard of Oz, or my 6-year-old self outing Santa. Truth be told, the "workshop" behind this part of the first e-Christmas is a low-rent facility in grungy Commerce, Calif., just off Interstate 5 south of Los Angeles.

It really was America's first Internet-driven Christmas. When all the mouse clicks and money get counted, consumers will have spent about $4 billion buying goods and services (including travel) online during the fourth quarter of 1998, and nearly $10 billion throughout the year, according to the Boston Consulting Group. Technology Scrooges may note that these sums are piddling--less than 1% of total U.S. retail sales. But in a good but not great pre-Christmas season, e-commerce sales more than tripled from the e-year before. With mouse and modem, consumers in more than eight million households voted that they'd rather buy online than fight for Furbys and parking places at stores.

This is the beginning of a real revolution. Not only is the Internet the fastest-growing medium in history, but online shopping is the most dramatic change ever to have hit retailing. It has caught on far more quickly than shopping centers in the 1950s, discount stores in the 1960s, and big-box category killers like Toys "R" Us and Home Depot in the 1980s. This holiday was the season when bricks-and-mortar retailers such as Brooks Brothers and Toys "R" Us reluctantly turned cyber--overcoming their "cannibalization inertia," or fear of pulling customers away from their stores. Any merchant that wasn't on the Web lost customers. America Online, the Internet's most popular mall, drew more than one million first-time shoppers in December; over the course of the holiday season, members spent an estimated $1.2 billion at stores that they reached through AOL. Amazon.com, the No. 1 online shopping site, sold so much merchandise in the quarter--CDs, videos, and gifts, as well as books--that the packages, placed end to end, would stretch more than 100 miles.

In many ways, eToys (www.etoys.com) is typical of the top Internet retailers. It's striving to be the Amazon of its industry. It's changing the model of how to sell. It's losing money by the millions. And it can afford to because so much venture capital--almost $2 billion in 1998, according to Venture One, a San Francisco research firm--is financing the birth of e-commerce.

That money is critical to e-tailing's future. For all the wonder of this business--you don't need capital to build stores or buy cash registers--the price of success is surprisingly steep. Hanging a shingle--setting up a pretty, easy-to-navigate Web page--is the least of it. Internet retailers must spend enormously to acquire customers and to master the intricacies of logistics and good service. Says Toby Lenk, eToys' founder and CEO: "I can't tell you how amazingly complex this business is. To be really good at this is really hard."

Before launching eToys in October 1997, the lanky, balding Harvard MBA had a successful career at another kind of virtual-reality company, Disney, where he rose to become head of strategic planning for the theme parks. But Lenk was bored, so he quit. Initially he tried to persuade venture capitalists to back a plan to sell high-priced toys to yuppie parents via the Internet. "That was a big yawn," says Michael Moritz of Sequoia Capital, which bankrolled such high-tech winners as Yahoo and Cisco Systems. Moritz and other venture capitalists pointed out to Lenk that Web merchants, more than traditional retailers, must offer very broad selections of products; e-shoppers demand it. So Lenk rethought his scheme and conceived eToys as a mass-market retailer that would carry a broader variety of toys--from Barbie to Brio--than any other store on earth. The venture capitalists loved this strategy; Sequoia and other firms have pumped more than $15 million into eToys.

Today the company has the attributes e-tailers need to succeed. It is tapping a huge market, the $22-billion-a-year toy industry. It was the first real Web player in the category--a major competitive advantage. Explains Lenk: "In land-based retailing, being first means you open up on one street corner. But on the Internet, you're a national player the day you open for business. You can become the de facto brand leader really quickly."

What's more, eToys is offering a solution to consumers' biggest complaints about shopping in stores: They don't have time to do it, and according to Boston Consulting Group's surveys, a sizable and growing minority downright loathes shopping. Moritz counts himself in that group. "The premise of eToys is that you never have to go to Toys "R" Us again," he says. "That's one of the most compelling investment premises I've ever heard."

Of course, being first and having a good strategy doesn't assure survival. The more difficult hurdle in Internet retailing these days is acquiring customers. Even Amazon, the e-tailing king, spends an estimated $29, on average, to acquire a new customer--which is why the company still isn't profitable. David Pecaut, who heads Boston Consulting Group's e-commerce group, says a typical cyber-retailer spends a staggering 65% of revenues on marketing and advertising. Marketing costs for bricks-and-mortar merchants, in contrast, typically run less than 5% of sales.

Even though eToys was the pioneer in toys, it faced competition from thousands of other Websites hawking wares. So Lenk had to ask, How do I get shoppers to my site? First off, he spent $3 million to become an "anchor tenant" on America Online; the deal gave eToys a presence and promotional space on AOL's Website for two years. He also gives 25% of each sales dollar to other Net inhabitants--5,000 of them, such as USA Today and Ameritech--whenever they steer a new customer eToys' way. Such revenue-sharing is standard practice in Web retailing, though in the conventional world it is akin to Barnes & Noble's paying a commission to a guy who stands on the street corner and directs traffic to its stores. "Some people thought we were crazy," says Lenk, whose 25% commission is unusually high. "But this business is all about getting customers. If we sell $40 worth of toys, $10 is a reasonable acquisition cost."

The costs mount. To become a big player, an e-commerce company must advertise, and not just on the Web. "Can you build up a business just by branding on the Internet? The answer is an absolutely resounding no," Lenk says. He learned this last fall, when eToys got a surprisingly large sales pop from its first print and TV campaigns. The ads carried the tag line "We bring the toy store to you"--a not-so-subtle dig at Toys "R" Us. The company also won a starring role in a Visa commercial--a boon, since consumers' No. 1 concern about shopping online is credit-card security.

So driven was eToys to draw shoppers that in the month before Christmas it handed out free the most sought-after toy of the season, Furby. Lenk couldn't get enough of the furry runts to sell, so he decided to stage an online sweepstakes and give away one per day. Through the promotion, eToys collected e-mail addresses from thousands of potential customers. This year the company will be zapping promotional pitches into their online mailboxes.

As in old-line retailing, once an online retailer brings in customers, the pressure is on to deliver. A surprise for startups like eToys is just how complex the logistics can be. Much of the difficulty relates to distribution. Some categories--books, music, computer hardware and software--are ideal for e-tailing because these industries have large, highly efficient distributors that ship overnight to merchants or even directly to customers. But the toy business, like many others, has no dominant distributors. That means Lenk must buy his merchandise from more than 500 manufacturers. And since shoppers--particularly e-shoppers--require selection and speedy service, eToys must carry lots of inventory, which is stacked sky-high in the stadium-sized warehouse. "The original notion of Internet retailing was that you never need to touch the product, you outsource everything, and you run an empire with hardly any people," the CEO says. "That's totally wrong. The virtual company simply doesn't work."

Now that the first e-Christmas is over, Lenk is exhaling with relief. He had projected that fourth-quarter sales would multiply 15 to 20 times vs. 1997. They far exceeded that; estimated full-year revenues were $15 million. Even so, Lenk was prepared. During the December crunch, he says, eToys shipped 95% of customer orders within 24 hours--a very good performance. It spent heavily to satisfy customers--giving free upgrades from regular to express shipping, for example, and $5 eToys coupons for customers who couldn't get toys because they were out of stock. These are smart moves. Studies show that a satisfied e-commerce shopper talks about the experience and draws six new customers to a merchant's site.

The company survived the season without major technical glitches--a claim many e-tailers, including Toys "R" Us, can't make. And eToys weathered the arrival online of Toys "R" Us just fine; the megamerchant's Website offered no more than half of eToys' merchandise and drew far fewer customers. Media Metrics, which tracks Internet traffic, says that eToys had 3.4 million visitors during the holiday season, almost three times as many as the Toys "R" Us Website. It even surpassed Barnes & Noble. "We just kicked butt," Lenk says.

For eToys and other startups, the challenge now is not just adding customers but getting existing ones to buy more. "That's the real trick of Internet retailing," says Sequoia's Moritz. The gold standard is Amazon, which derives almost two-thirds of its sales from repeat customers. The key to getting repeat business, of course, is terrific customer satisfaction.

Aspiring to this, Lenk says he will probably double his merchandise selection this year and move into new categories--maybe kids' sporting goods. He'll start shipping internationally. He'll open an East Coast warehouse, because 70% of e-commerce shoppers live east of the Mississippi. And he will work on improving his seemingly low-tech warehouse in California. "Our warehouse is similar to Amazon's. It's sophisticated for an Internet retailer," Lenk says. What you can't see when you're inside is the computer systems that track eToys' inventory and handle orders. Still, the facility is nowhere near as automated as the warehouses of premier direct sellers such as L.L. Bean. "Hopefully, we'll get to that level someday," says Lenk, speaking for the virgin e-commerce industry.

The success of eToys, he insists, depends on execution. "If we don't deliver what customers expect, somebody will beat us. Maybe Toys "R" Us. Maybe Wal-Mart." Or maybe Amazon, which offered 185 children's toys this Christmas. As eToys expands its merchandise and as Amazon moves aggressively beyond books, the two will probably become more direct rivals.

A key question for eToys and all these growth-obsessed Internet players would seem to be, When will they make money? Lenk dons his old MBA hat and talks around the issue: "We've certainly given thought to an endgame business model." Which means, in English, that eToys plans to earn a profit someday. When? Lenk won't offer even a hint.

For now, the game is to spend, spend, and spend to acquire customers. Venture capital money isn't adequate, so eToys this year is expected to do what every other hot Web company is doing: an IPO. That will provide cash to build the business further. Eventually, eToys will have so many customers--loyal customers--that it will no longer need to spend as lavishly on marketing. Then profits will come. At least that is what's supposed to happen. There is no history here. We'll need to wait a few more e-Christmases to find out.