When Garden.com went public on September 16, 1999, its workers danced in the halls. They had been glued to their computer screens throughout the day, watching with a sort of dreamy, unalloyed happiness as the price of their shares floated upward. The online gardening retailer’s stock had opened at $16.60, climbed briefly to $21.25, and then settled back to $19.06. When the market closed, the company’s three founders—Cliff Sharples, his wife, Lisa, and Jamie O’Neill—were each worth $2 million and change; their employees had watched their personal wealth expand dramatically in a matter of hours. And now that they were all well on their way to being rich, the celebration began. A deejay blasted rock and roll from the back of the company’s headquarters, a converted warehouse just off Mopac in North Austin. Kegs were broken out. Cases of champagne were brought in, and a massive, happy party raged on into the night. The whole thing seemed just short of miraculous—all the more so since the three-and-a-half-year-old Garden.com was known to be one of the best companies in town to work for, with a tolerant, loosey-goosey culture that allowed people to walk around barefoot and bring their dogs to work. Its symbol was a giant carrot suspended from the ceiling near the front door. To the company’s 169 employees, the success of the IPO, which gathered $54 million in cash that day and created a number of paper millionaires, was nothing less than cosmic justice.Sixteen months later the carrot hung in an empty, echoing space where furniture and telephones were stacked in piles, and lawyers and other vultures were picking through the ruins. On November 15, 2000, Garden.com had announced that it had been unable to find more money to keep itself alive and therefore was going out of business. By December 31 it was just another piece of e-commerce roadkill on the information superhighway, an apparent example of capitalism’s capacity for wretched excess.

What’s interesting about Garden.com is not that it was a company that tried and failed to make money by selling things on the Internet. More than 130 Internet companies have gone out of business since January 2000, and many more will follow. What is compelling is that the founders seem not to have made any of the mistakes that so many other dot-coms have made. In fact, in all ways except the most important one—staying solvent—they succeeded brilliantly. They did not, like now-defunct clothing e-tailer Boo.com, pay out $100,000 allowances to the principals so that they could decorate their homes, or fly executives on the Concorde, or spend $42 million in eighteen months on largely useless advertising and promotion. They did not, like now-defunct pet product e-tailer Pets.com, foolishly sell goods for less than they cost in an effort to lock down market share. Their Web site technology did not fail them. They did not buy Porsches. Indeed, Garden.com met or beat every single economic performance target that had been set at the time of its public stock offering. Its managers did exactly what they said they’d do, exactly what investors had given them $105 million to do, and exactly what industry analysts had said they needed to do. And still they went down. As original investor John Thornton of Austin Ventures, a venture capital firm, puts it, “This was the nicest, most talented bunch of people I ever lost all my money with.”

So why did Garden.com die on the vine? The answer lies in the nature of the company that the Sharpleses and O’Neill thought up in the garage of the Sharpleses’ house five years ago. The three had gone to business school together at Northwestern University’s Kellogg School and had all landed in Austin on the same day in May 1995, working for software maker Trilogy. Ten weeks later they quit their jobs to start a new company. The only problem was, they had no idea what sort of company. “Cliff and I had just gotten married,” says Lisa. “We had moved here, bought a house, and hadn’t even made our first mortgage payment yet. My parents thought we were nuts. My grandmother sent me a letter with cash in it because she was worried about us.”

Though none of them were avid gardeners, they quickly zeroed in on gardening as the place to build their e-business. It was a $50 billion industry that had no dominant companies and no one at all selling gardening products over the Internet. This was in the earliest days of e-tailing; Amazon.com had gone into business only a few months before. Subsisting on O’Neill’s Visa card and peanut-butter-and-jelly sandwiches, they came up with a breathtakingly ambitious plan to create a $75 million company that would hold almost no inventory but instead would use a network of suppliers to deliver thousands of gardening and related products to consumers. It would be a so-called pure-play e-tailer, meaning that, unlike a bricks-and-mortar store such as Home Depot or a catalog company such as Lands’ End—both of which do business on the Web—the company’s business would be built entirely on the Internet. “The most compelling thing was that there wasn’t some big guy who could come and crush you like a bug,” says Lisa. The goal: build the first national brand of gardening that people recognized. Cliff became the chief executive officer, Jamie the chief operating officer, and Lisa the head of marketing.

The company got its first $750,000 from Austin Ventures in December 1995 and opened its Web site for business in March 1996. Fueled by several more rounds of financing, Garden.com grew quickly. It raised $2 million more in 1996, $5.25 million in 1997, then major injections of $20 million in 1998 and $22.6 million in 1999. By then it had built what many considered one of the best commercial Web sites on the Internet. Business boomed. Garden.com now had some seventy suppliers across the country and was selling all sorts of things in addition to azaleas, tulips, hydrangeas, and roses, including furniture, candles, lamps, shoes, soap, tea, cookware, perfume, peppermint foot lotion, and Christmas tree ornaments.

The company had also begun to acquire a reputation as both a progressive, compassionate employer and a first-class e-tailer. “This was definitely a different company,” says Cecilia Nasti, who worked there as a “garden doctor,” answering gardening questions and writing online care guides. “I barely wore shoes. I walked around barefoot. People’s dogs would come back and visit you at your desk. I know this sounds strange, but there really was a lot of love in the room. There were a lot of good friendships, deeper than the usual work friendships.”

Garden.com was white-hot. Texas Monthly Biz trumpeted its success in two articles, and both Northwestern’s and the University of Texas at Austin’s business schools used the company as a case study for their students. In a lengthy profile in Inc. magazine in the summer of 1999, just after the company had raised $22.6 million more in private capital, the writer observed that “investors are pounding down the door, coughing up more money than the company knows what to do with.” One of the company’s directors, Steven Dietz of Global Retail Partners, was quoted as saying, “I kept arguing we didn’t need more than twelve million dollars. But there were a lot of good, quality people who wanted in on the deal.” The illogic of that didn’t seem to make any difference. Nor did the fact that, in spite of more than $50 million in investments, Garden.com had only $5.4 million in revenues to show for it and a loss of $19 million in the fiscal year that ended June 30, 1999. In September the company went public anyway.

That was also the month that the investment community collectively took leave of its senses, at least as far as e-commerce was concerned. “In the fourth quarter of 1999 everything was in a complete frenzy,” says O’Neill. “It was euphoria. In the third and fourth quarter a lot of companies raised tons of money.” For Garden.com, however, this would soon present a major problem. The company had been relatively conservative in raising money and relatively stingy with salaries too: Cliff, Lisa, and Jamie made $113,679 each, and the company’s top technical officer was making $143,000. But now other companies were raging out of control. “We lived for two and a half years on $8 million,” says O’Neill. “A lot of our e-tailer brethren had a first round of $10 million and a second round of $50 million. People looked at that and expected a sort of vertical growth rate for these companies. That didn’t happen, so they started saying that these companies, and the industry in general, might be mortal. And that was when the blood went into the water.”

The high-tech meltdown in the Nasdaq market started in April. For Garden.com, which needed to raise yet another round of private capital, it was a spectacular piece of bad luck. Suddenly Cliff Sharples and Jamie O’Neill, who had had nothing but good luck raising money, were getting polite no-thank-yous. They hit up everybody they could think of, looking for investors or partners or potential purchasers. “We saw a stronger and stronger wholesale rejection of the category,” says Cliff. “It was almost a fear of e-tailers, a fear of investment. We heard this message pretty consistently.” As spring bled into summer, the market plunged farther, dragging Garden.com’s shares down too.

All this was happening in spite of the fact that the company was performing precisely as it was supposed to. Its award- winning Web site had signed up 1.5 million members, including 290,000 paying customers. The company now employed some 275 people. For the fiscal year ending June 30, 2000, it had $15 million in revenue and a loss of $38 million, exactly the performance that industry analysts had predicted when the company went public. The next year’s revenues were expected to hit the low forties, with a much smaller loss. The year after that, according to its financial analysts, the company stood a good chance of breaking even.

It would never get the opportunity. In June came the first layoff—24 people. Uncharacteristically for such a touchy-feely culture, the terminated employees had their computer access shut off and were escorted to the door by fellow employees. The principals would not make that mistake again. In September, 96 people were laid off, more gently this time. Cliff and Jamie redrew their business plans, desperately trying to cut costs, but to no avail. Investors were no longer putting even a penny into what were now perceived as the cash-sucking black holes of e-commerce.

If you listen to Cliff Sharples, Garden.com died because of market psychology and bad timing. But according to Austin Ventures’ John Thornton, a member of the board who owned 8.6 percent of the company when it went public, the problem ran far deeper. “Their model was just not viable,” he says. “We learned that eventually it’s just too damn expensive to get people into your store. That’s the whole story.” According to Thornton, the basic fallacy of all of the pure-play e-tail start-ups was that somehow millions of people were going to come to an Internet site the way they might walk into a store. That did not happen. To get customers to their sites, e-tailers had to spend a fortune on advertising. Garden.com, in fact, spent 56.7 percent of its IPO proceeds on various forms of advertising and marketing, including hugely expensive glossy catalogs and even a full-scale magazine. “They did not understand what it would cost to attract the traffic,” says Thornton. “That’s the story. You can boil it all down to that. This was a completely new world. Nobody had a clue what it was going to cost, and we and everybody else were wrong about it.” Cliff Sharples concedes that the costs to advertise online were “prohibitively expensive.” But he argues that by using direct mail and catalogs the company had driven the cost of getting new customers down. “It cost us sixty dollars to get a customer in spring 1999,” he says. “By fall 2000 we were down to twenty to thirty dollars.”

On a windswept day in December, the giant carrot was still suspended above Garden.com’s vast sea of cubicles in North Austin. But almost everyone was gone. In keeping with the company’s culture, Cliff and Lisa and Jamie are running an orderly shutdown. Employees get severance pay. Because management had made a decision early not to run the company into bankruptcy, they appear to have enough cash to pay off their debts. “After spending five years and pouring your heart and soul into it,” says Cliff, “we didn’t want to be reduced to a story of these idiots who ran out of cash and then just walked away and left a big mess for the lawyers to sort out. That’s not the epitaph we want.”