(FORTUNE Magazine) – The Gartner Group has always billed itself as "the world's leading authority on information technology." That's a weighty claim, but it's nothing compared with what fans of Gartner's stock once said about the Stamford, Conn., research firm. Two and a half years ago, with its shares at $27, analysts on Wall Street lined up to tout Gartner as the way to play the tech revolution if you were afraid of investing directly in volatile tech stocks. The company looked like a sure thing, poised to deliver 35% earnings growth quarter after quarter. Even its ticker symbol--IT--seemed perfect.

Investment bank Alex. Brown went so far as to predict that Gartner's shares could hit $44 within a year. They never came close. Instead, the stock slid through most of 1998, and by this fall was down to $20. Then, in late October, Gartner missed fourth-quarter earnings estimates and warned that new spending on e-business would mean drastically lower earnings in 2000 as well. By mid-November, the stock had sunk to $10. Meanwhile, other information-technology research firms suffered similar fates. While Forrester Research and Jupiter Communications have fared well, Meta Group is trading far below its high for the year, and Giga is at an all-time low.

The performance of these companies is a cautionary tale for anyone investing in technology on the periphery, betting on companies one step removed from the sector in an attempt to get returns without risk. In tech, there is always risk, even on the periphery, and what happened at Gartner is a prime example. It is the story of how the world's leading authority on information technology mishandled the decade's biggest tech trend, the Internet.

First, a little history. Gartner, which has been around since 1979, generates more than half its $734 million in annual revenues from research, providing analysis of tech trends and products to corporations that use the information to decide which computers and software to buy. Gartner has about 10,000 customers, many of whom spend hundreds of thousands of dollars each year to subscribe to its reports. The rest of Gartner's money comes from a burgeoning consulting arm and from hosting popular, not to mention lucrative, tech conferences.

This business model has allowed Gartner to more than triple its revenues since 1993 and generate fat profit margins. Unfortunately, this success has also bred complacency. Since the mid-1990s, annual earnings growth has slowed from 75% to 10%. "Like many tech companies, the success of their initial model made it hard to move to the next generation," says CS First Boston analyst Mark Wolfenberger. "Gartner's product line has not kept pace with what its customers are looking for."

What customers are looking for these days is advice about e-commerce. And while Gartner analysts wrote about the Net's business potential as early as 1993, the company itself hasn't adapted to the changes wrought by the Net. Most of its researchers remain focused on traditional infotech areas like data storage and the relative merits of Unix vs. WindowsNT. Given the pace of change on the Internet, such sloth is a particularly deadly sin.

Some observers add that Gartner's biggest blunder was attempting to address the e-business market through a minority investment in upstart Jupiter Communications in October 1997. Gartner eventually hoped to purchase all of Jupiter. But Jupiter, which quickly established itself as a top e-commerce consulting firm, had other ideas. Its board rejected Gartner's offer to buy the company outright and decided to go public this fall. That made Gartner a happy investor (it still owns 28% of Jupiter post-IPO) but left it with a gaping hole in its research offerings. Gartner belatedly launched its own e-commerce product line, GartnerE, this summer, but analysts remain skeptical. "They're definitely late," says Manoj Mittal, an analyst at BancBoston Robertson Stephens. "But the market opportunity is big."

The Internet poses other challenges for Gartner too. For one thing, it has put pressure on the price of Gartner's research, since the basic market analysis for which the company once charged thousands of dollars is increasingly available free online from a variety of sources. Thirty-four-year-old Michael Fleisher, who took over as Gartner's CEO in October, wants to ensure the research firm can respond to these Internet threats. He plans a major push into business-to-business e-commerce, hiring 100 additional e-business experts to bolster the company's existing 245, and increasing its current 400-person consulting force by 50%.

None of this will be easy--or cheap. Gartner will spend an additional $50 million to $60 million next year to implement its strategy, most of it on hiring, training, and compensating workers. That's the reason the company says its earnings growth will be depressed next year. "We believe business-to-business e-commerce is about to explode in a spectacular fashion," Fleisher says. "We want to have the same dominance in this new marketplace as we have always had in the infotech research field." So far, analysts like Fleisher's plan but say timing Gartner's recovery is tricky. "Sixty million is a pretty substantial investment, and it remains to be seen if they can execute their strategy," says Sandra Notardonato, an analyst at Adams Harkness in Boston.

True, at its current price, Gartner might be attractive to bottom-fishing investors. But Kevin Landis--the well-respected money manager at San Jose's Firsthand Funds and a veteran of Dataquest (now owned by Gartner)--still thinks that Gartner and the entire research sector may be overvalued. "High-tech research is the art of forecasting what hasn't happened yet," he says. "A lot of it comes close to educated guessing. And if you think people don't want to pay for most types of information these days, wait to see what they will pay for educated guessing."

PLUS: GARRETT VAN WAGONER'S COMEBACK | QUALCOMM'S AMAZING RISE | CACHE IS KING