Sunday, May 23, 1999

The power behind portals

Inktomi searches open most Internet doors

USA Today

        SAN MATEO, Calif. — Dave Peterschmidt doesn't care that Inktomi isn't a household name. In fact the chief executive officer of the Internet software maker says he doesn't really care whether Main Street America knows about his company at all, as long as potential customers and competitors do.

        That's because Inktomi's goal is not to be a centralized directory of what's on the Web, like so many other companies that started out as search engines. Instead, it's out to be the software engine that powers those portals.

        If Inktomi were a man, he'd be wearing overalls smeared with grease.

        “Our business model is really predicated on not being a portal at all,” co-founder Paul Gauthier said.

        He was a graduate student at the University of California when he started the company in 1996 with associate professor Eric Brewer.

        Using a concept known as parallel or cluster computing, the two researchers made breakthroughs in linking computers to create the effect of a larger, more powerful computer that could handle the massive information-retrieval problems presented by the Internet.

        The two started working on a search engine, which they named Inktomi after a trickster spider in Lakota Indian legend who defeated others through wit and cunning.

        The prototype search engine quickly became popular among nerds and Net aficionados, who in the old days had few places to go for searches.

        The two men brought Mr. Peterschmidt in to help manage their fledgling operation.

        Inktomi now offers searches, a shopping engine and traffic-management services for Internet service providers. Its client list includes America Online and portals, such as Yahoo and Snap, which many people use as starting blocks for finding information on the Web. While most portals put together their own directories, they rely on specialized software to pull up results from the Internet. That software is often Inktomi though some competitors, including Excite, Lycos and Compaq's AltaVista, have developed their own.

        Inktomi said its advantage rests in its noncompetitive stance, supplying others with the technology they need to run their sites.

        Analysts agree.

        “Being Switzerland is pretty central to our model,” Mr. Gauthier said.

        Judging by Inktomi's customer list and its stock price — which has soared from $15.371/2 when it went public June 10, 1998, to a high of $159.121/2 April 12 — it is doing well.

        Some analysts aren't convinced that powering other companies' Web sites is enough. Said Internet analyst Abhishek Gami with William Blair & Co.: “It makes me a little bit nervous that they have a limited product portfolio.”

        Others have applauded Inktomi's move to stay at its core business of being an infrastructure player but note that it faces competition in all arenas.

        Goldman Sachs analyst Rakesh Sood said Inktomi could face trouble “if niche competitors in any of these given areas target them.”

        Andrew de Vries, spokesman for HotBot, the first search engine to use Inktomi, lauded the company's technology. But he said it is not a panacea.

        Like all businesses competing on Internet time, success depends largely on risk-taking — balancing current work against developing the latest innovation. That's not always easy.

        Mr. Peterschmidt likes to tell a story about how he almost killed the company.

        Right after he joined, he took the employees — all 13 of them — on a retreat to Monterey, Calif., a thinking kind of place with lots of trees and golf courses.

        “We were sitting in the strategic planning session. I started the session by saying, what do you want to do with Inktomi? And they all said, we want to make money. And I said do you want to make a little money or a lot of money. And they said a lot of money.”

        Inktomi at that point was just starting to squeeze out some cash from its search engine.

        But to be really big, it had to immediately change from being a search engine company to an “Internet infrastructure company,” as Mr. Peterschmidt put it. So he decided to split the tiny company's resources and tackle the next major problem of the Net: helping traffic flow over the increasingly congested Web.

        “I'm driving back, and I'm thinking I may have made a decision that could kill the company. I could have taken critical mass away from our only source of revenue.”

        Of course, the very fact that Mr. Peterschmidt can talk about it with a smile from his small office here in the Silicon Valley, is proof the strategy has worked. So far.

        “As the Internet gets larger, we continue to make it faster and more robust,” Mr. Peterschmidt said.

        This year there are an estimated 400 million Web pages on the Internet, up from fewer than 30 million in 1992.

        He predicted that by 2002, there will be 8 billion Web pages.

        People will be relying on search engines more than ever to find what they need, he said. “We're hitting an inflection point that's like nothing anyone's ever seen.”


Blue Chip Venture Co. ascends into cyberspace
Wanted: Drug testers
Companies brought to to NxLevel
Chamber lauds ingenuity, service
Tech firms courting senior citizens
Ever-beloved typewriters hang on
- The power behind portals