SAN FRANCISCO -- Life couldn't be much sweeter right now for search-engine company Ask Jeeves (ASKJ) .

Wall Street's infatuation with the Berkeley, Calif., company resulted in the third-largest one-day gain for an Internet initial public offering. And by the end of its second day of trading, the stock jumped even further to 72, or 414% above its offering price of 14. The company burst into the Fourth of July weekend with a valuation of $1.75 billion. Not bad for an outfit that lost $4.9 million on revenue of $1.1 million in the first quarter of 1999.

The press dutifully responded with a chorus of cheerleading. "Ask Jeeves: Net Stocks Not Dead," declared the

New York Post

. "Ask Jeeves ... Hits It Big," blared the

San Francisco Chronicle

. Even

The Wall Street Journal's

interactive edition ran a sycophantic piece detailing the sizable paper gains pulled in by the company's investors and executives.

But looking under the hood of the company, it's possible that the Street's infatuation with Ask Jeeves could end in heartbreak. That's because the key part of its business model -- the part that is rocketing its valuations into the stratosphere -- is a totally unproven service in a very crowded market that even the company admits it has doubts about.

"There is a tactical need for a product like Ask Jeeves, but is it a sustainable business?" asks Carter Lusher, a research analyst with the

Gartner Group

. "It's not like they have total exclusion in the market."

According to the breathless accounts of Ask Jeeves' IPO, investors were stoked about the company's search engine, which allows people to ask simple questions such as "

What is the most overhyped Internet company of the year?" Compared with the rest of the search engines that operate through arcane and often unreliable keyword searches, the company fronted by a cheesy butler has a unique story.

But if you ask Christopher Lord, portfolio manager for

Pivotal Asset Management

, a firm that invests in high-tech companies that have just gone or are about to go public, the real story about Ask Jeeves isn't its advertising-based search engine. It's its corporate question-answering service.

In a nutshell, the corporate question service is a virtual customer-service representative. It's the latest example of what analysts term "customer relationship management," which has traditionally been dominated by call centers and, more recently, by email. Although the service only accounted for 6% of the company's revenue in the first quarter of 1999, up from 3% in all of 1998, Lord says the market is potentially enormous.

"The idea is to become embedded in a corporation's Web site and bring in recurring revenues," explains Lord, who also is a research analyst for

Amerindo Investment Advisors

. "It's a very powerful business model." Amerindo, says Lord, will probably invest in Ask Jeeves, too.

Robert Wrubel, the 38-year-old CEO and president of Ask Jeeves, says the corporate question service has around five to 10 customers, including


(DELL) - Get Report








Office Depot

(ODP) - Get Report

. As companies find a need to interface with customers on the Web, the technology is promising, he says.

But the service is no sure-fire slam dunk. To begin with, Ask Jeeves admits in its prospectus that "we cannot yet determine the effectiveness of our Corporate Question Answering Service compared to traditional methods of customer relationship management." Far from the typical risk factor, Ask Jeeves seems unsure that its most promising product line will even catch on.

What's more, even if the service is useful, it must compete against a host of products from such customer-management software companies as





(INFR) - Get Report







, some of which may offer a broader-based solution.

"Inference or Primus can take in knowledge from a variety of different places and index it," says Gartner analyst Lusher. Both of them can index other information sources such as documents from


(MSFT) - Get Report

Word and


Notes, as well as knowledge from customer-service representatives. By contrast, Lusher says the current version of Ask Jeeves can only index Web-based information in HTML, the computer language of the Web.

Asked how his company can compete with bigger customer management software rivals offering a broader array of products, CEO Wrubel says they are more likely to end up as partners of Ask Jeeves than competitors.

The outlook is similar for Ask Jeeves' search engine, which analysts say is a good idea but not a panacea. "There are a lot of tools out there that are going to be fighting in this space," says Cormac Foster, an analyst with

Jupiter Communications

, who notes that bigger search companies like






could end up as potential competitors. "By no means does Ask Jeeves have a lock on everything, but they have a head start on a consumer-friendly interface with a natural-language processing engine."

Ironically, one of Ask Jeeves' supposed competitive advantages, its staff of 100 human editors, may be an Achilles' heel. "They need to demonstrate that their editorial model doesn't become a bottleneck," says Lusher, who notes that competing services may offer quicker, more automated solutions.

Similarly, Ask Jeeves may have stumbled by not branding its corporate question service with the Ask Jeeves' brand -- one of the company's key assets. For its contract with Dell, the company named the service Ask Dudley.

"I think that's a big mistake," says Steve Harmon, senior investment analyst with


. "At some point the brand is diluted. It's like Coke saying you could buy their syrup and rename it Joe's Coke. They ought to perpetuate the brand. It just means they need to spend more money."

One thing is clear, however: Within the customer relationship management market, Ask Jeeves is trading at an Everest-like premium compared with its more earthbound rivals. Ask Jeeves' market-cap-to-1998-revenue multiple is 2,666 times. That compares to 0.88 for Inference, 1.94 for Vantive and 7.34 for Clarify. But even using analysts' 1999 revenue estimates for Ask Jeeves, which are pegged at around $12.5 million, the company's multiple is still a high 126.4.

"I think that paying more than 500 times revenue you have to wonder if they're going to scale fast enough," says Harmon. "The strength is in the concept but it hasn't shown the potential it can offer."

So much for the death of the hypervalued Internet IPO.