Yahoo looks to improve search experience

September 30, 2009 By Mike Swift

When Yahoo CEO Carol Bartz announced a $100 million global brand campaign in New York recently as the company launched its new Internet search "experience," research scientists like Preston McAfee, Duncan Watts and David Reiley were not in the spotlight.

Yet the three scientists, who recently quit tenured posts at leading U.S. universities to join Yahoo Labs, are a crucial part of Yahoo as it tries to improve the flagging popularity of its search engine and attract audiences to its other products, including e-mail, instant messaging and software for mobile devices.

In a news conference at the Nasdaq headquarters to announce the multi-platform "It's You" campaign that will play in the United States and nine other countries around the world over the next 15 months, Bartz said Yahoo believes it can build its search audience by focusing on the user "experience," rather than viewing search as a raw data query.

A basic Web search, Bartz told reporters, is like an Intel chip -- a common ingredient many computer makers use to create their distinct products.

"The experience that H-P wraps around those chips is different than the experience Dell wraps around those chips," Bartz said. "We are looking to focus our research and our scientists on the experience around that basic information. So search is vital for us."

Yahoo's decision to focus on the "front-end" human experience of search provides a window into its strategy in its heavily criticized search partnership with Microsoft, which was panned by investors and many analysts when the deal did not include a hefty upfront payment from Microsoft. And what's notable about many of the stable of researchers at Yahoo backing that effort is that many of them, including McAfee, Watts and Reiley, aren't even computer scientists.

They are part of a team of social scientists -- cognitive psychologists, sociologists, economists and ethnographers -- that Yahoo hopes will help close the search gap with the dominant Google. Analysts say that effort is crucial to Yahoo's future.

"You have to fix search," said Karsten Weide, an analyst with the research firm IDC who was initially critical of Yahoo's partnership with Microsoft.

The plan to broaden Yahoo Labs into a multidisciplinary team where social scientists work directly with computer scientists is one element of Yahoo's strategy to hold people on its Web properties, after its new branding campaign -- Yahoo's single largest integrated global campaign ever -- brings them in the door.

"What we want to create over the next 10 to 20 years is that science of understanding of what makes a Web experience 'sticky,'" said Prabhakar Raghavan, Yahoo's senior vice president for search strategy and the head of Yahoo Labs.

Yahoo's social scientists, Raghavan said, "are people who care about what it is that makes an engaged audience -- Why do people hang out on our front page? Why do people search? Why do they go to Facebook?"

The improvement in search functions is part of an overall upgrade, including recently unveiled updates to Yahoo's core e-mail and instant messaging products, as well as revisions to its home page.

Raghavan said the company is betting that improving the experience of search -- getting better at producing the information people intuitively want when they type a keyword, showing it to them in a way they want to see it, and proving that advertising on Yahoo has a direct connection to improved sales -- is the way to close the gap with Google.

That is part of the philosophy behind Yahoo's new search results page, whose layout and content is partly based on research such as where the human eye tends to linger on a Web page. A search for "San Jose Sharks" Tuesday on Yahoo's new interface produced not just the familiar 10 hyperlinks in the center of the page, but a box on the left-hand of the page with icons for ESPN, Wikipedia and other sites with Sharks content, as well as a box of related topics such as the NHL team's new jerseys.

"We want to take all the smart people and all the good machines that have been churning away on the back end, and say, 'You know what? Let's think front end,'" Raghavan said. "Now, is that a winning strategy? We hope so; we believe so, but that is the embedded viewpoint that we are building on."

The viewpoint is behind Yahoo's search partnership with Microsoft, awaiting government approval. While Microsoft would handle the back-end hardware and software and its new Bing search engine would provide the underlying search results, Yahoo would pocket 88 percent of the revenue from searches done from its sites.

While Bartz and other Yahoo executives faced skeptical questions about the company's strategy in New York, some former critics say they like the front-end strategy.

"I'm now saying this is a smart thing to do," Weide said of Yahoo's decision to leave the expensive investment of updating the back-end of search to Microsoft. "In terms of the search engine interface with the user, there is a lot of room for improvement."

Yahoo's principal scientists are expected to be academic leaders, publishing papers on their research just as university professors do, while also working on direct business problems for the company.

Watts, a sociologist who recently resigned professorship at Columbia University to work for Yahoo Labs, said he was drawn partly by the chance to work directly with scientists from other disciplines, something that he said doesn't happen that often at many universities.

Yahoo Labs "is more like academia than academia," Watts said, and is very different from other corporate labs, where there are not a variety of scientists from different disciplines and there is a more exclusive focus on solving specific business problems. "Google doesn't do this."

Economists such as Reiley and McAfee, who recently resigned tenured jobs at the University of Arizona and the California Institute of Technology, respectively, say they were drawn by the chance to tap Yahoo's rich data about its 581 million monthly users, and the chance to work on real-world problems.

"Economics has been historically a mostly theoretical science and very abstract -- not an applied science at all," McAfee said. "I'm part of a movement that has an engineering model for economics, that is to say, we can build stuff."
___

(c) 2009, San Jose Mercury News (San Jose, Calif.).
Visit Mercury Center, the World Wide Web site of the Mercury News, at http://www.bayarea.com/mld/mercurynews
Distributed by McClatchy-Tribune Information Services.