Microsoft Researcher Questions Search Engine Business Model

Speaking before about 100 gathered for the Emerging Technologies Conference at MIT in Cambridge, Mass., Eric Brill, a senior researcher at Microsoft Research, said the "monetization" model of web search technology is at risk.

Brill's comments come on the heels of Google's highly successful IPO last month and bullish Wall Street reports this week that sent Google shares soaring to a high of $127.

"There's tons and tons of money to be made," said Brill. He acknowledged that Microsoft is intent on playing catch up with its own web search engine later this year yet projected that both companies could hit a fiscal wall. "There are two fears here: The better we get at search, the cheaper the ads are," Brill said. "Another possible problem is it is cheaper and cheaper to build a search engine. It could at some point get commoditized."

His brief -- and targeted -- comments annoyed one ISV who sat on the M.I.T. panel, entitled Next-Generation Search.

"I have to disagree with Microsoft saying search engine is not a formidable task," said Liest Capper, CEO of Australian ISV Mooter, which developed a web search engine that tracks down data based on sophisticated user profiling technology.

The panel included representatives from Mooter and other web search innovators including Blinkx, Nexidia and Dipsie. Blinkx, of London and San Francisco, offers an IE plug-in that provides contextual searches of data stored on local drives and the web.

Nexidia, Atlanta, offers high speed phoenetic searching that is capable of searching data far faster and more accurately than other engines today, company executives said.

Dipsie is an ISV in Chicago whose web crawling and indexing engine will search well beyond the number of web pages Google and Yahoo index today, said Jason Wiener, of Dipsie. He estimated that those two leading search engines scour only 1 percent of all the web pages available on the Internet.

Dipsie's SEO offering will be announced to the market in the coming weeks, he said.

Brill had little to say about Microsoft's next-generation web search technology except that the traditional text-based hunt and peck search would vanish within three years. "What's wrong is we have document-centric views" of search results, he said. "We're trying to take a more info-centric view of search." The Microsoft researcher, however, avoided a question by the panel moderator on Microsoft's search technology plans for Longhorn.

Microsoft recently acknowledged that it will delay the release of the next generation file system in Longhorn code-named WinFS and will instead offer enhanced and faster searching in the Windows client upgrade due in 2006.

Recently, Microsoft CEO Steve Ballmer acknowledged that web search -- and corporate data search -- is a big focus for Microsoft. At another event in the Boston area last month, Ballmer pointed to AltaVista as Search Version 1.0, Yahoo as Search Version 2.0 and Google as Search Version 3.0. Microsoft, he said, will develop a next generation of search technology that would not only allow companies to search the Internet more effectively but search their computer networks, hard drives and e-mail more effectively.