For an Internet company, research is the creation of innovation -- the ability to plant seeds for growth several years down the road. And investors who don't know the nuances of search algorithms can still handicap a stock's long-term prospects by looking at the intellectual brainpower going into its research initiatives.

In the horse race between Google ( GOOG) and Yahoo! ( YHOO), the conventional wisdom is that Google has the edge in research: Its stock is up 240% since its IPO nearly a year ago, vs. a 26% appreciation in Yahoo!. Google is very finicky about its hires, which means if you get in, your job becomes a kind of daily Mensa party. And then there's the freedom -- if not the expectation -- to spend 20% of your time on a labor of love.

But it's far too early to write off Yahoo!, which has been steadily, if quietly, building a stable of brilliant researchers on its own. Under the leadership of Usama Fayad, Yahoo!'s chief data officer, the company has brought in its share of intellectual caliber, most recently in the hiring of Prabhakar Raghavan, head of Yahoo! Research, as well as a research partnership announced with the University of California, Berkeley.

So, with all of the Google hype, why would any brainiac go work for Yahoo!? Raghavan and others at Yahoo! offer two compelling reasons: First, Yahoo! has the Rosetta Stone of online customer data. For 10 years, it's collected the daily search, email, financial and entertainment habits of users, which number in the hundreds of millions today. And second, by extension, that data open up a lot of areas of research beyond search -- including fast-developing areas such as social networking, user interface and data mining.

"There's no bigger collection of data on this planet. We're collecting 10 terabytes of data a day, and that doesn't include Web pages and 'html' content, which may be the primary focus of someone at Google," says Raghavan, a research veteran from Verity and IBM. Raghavan is also a Stanford consulting professor and the editor-in-chief of the Journal of the Association for Computing Machinery.

"There's a tremendous opportunity here to address the mass consumer that has never really happened before in history," he says. "We get to study data and networking effects that show how people catch on to trends. And we get to study how to monetize all this, so there's an economic angle as well."

So, while Google workers may be feasting on arugula with dried apricots and capelin pesto, Yahoo!'s researchers are pigging out on deep data showing how online communities are built, how consumers open themselves to the long tail of entertainment niches and why sensible people would create and publish content without being paid a cent for it. Those questions are tightly woven with search technology, but extend well beyond it.

As a result, Yahoo!'s research, despite its relatively lower profile to date, is catching the eye of some on Wall Street, who see its potential leverage in future growth.

"Search is Google's sweet spot, but Yahoo! has a relationship with its customers and Google does not," says Scott Devitt, an analyst at Legg Mason, which has no underwriting relationship with either company. "I rate Yahoo! buy and Google hold, and the most relevant reason is the relationship and data that Yahoo! has with its customer base."

Yahoo! is also trying to get another edge by building stronger relationships with academic institutions like Berkeley. Whereas Google has longstanding ties with Stanford and contacts at a variety of top universities, Yahoo!'s making deeper inroads into academia with formal relationships.

In a nondescript office building in downtown Berkeley, a quarter-mile from the outskirts of the University of California, Yahoo! is starting an experiment that may have a significant impact on its future strategy. It's partnered with Professor Marc Davis, who came to Berkeley from MIT's pioneering Media Lab, to collaborate on work that Davis began in his Garage Cinema project.

At Berkeley, Davis developed a software for camera phones that can anticipate who you'd share a photo with based on who is in the photo. Bluetooth technology says where you are located and, most interestingly, what time of day the photo is taken. The project not only dovetails with some of the research Yahoo! has been doing in mobile media, but it can also benefit from Yahoo!'s ocean of customer data.

"One thing we've been looking at is how mobile devices can transform the way people work with media content," says Davis. "As a researcher, what I see in Yahoo! is hundreds of millions of people working together. I couldn't build that data set at Berkeley. Having access to what people look for and the media they create as a living, breathing community, that's an incredible draw."

Davis is already working closely with Yahoo! properties like Flickr, one of the earliest success stories in social networking. From a research standpoint, Flickr is something of a puzzle: How did a photo-sharing site develop a fiercely loyal community around it when Friendster, with all of its money and deep finance, couldn't?

By watching how Flickr's community functions and grows, Davis can try to tackle an even bigger question: How can Flickr's niche-community DNA be replicated on a vastly larger scale -- say a community of 200 million global users? Yahoo! is betting research will help it create such a community, not simply react to them after they are built elsewhere.

"Being able to access what people look for online will support that task," says Davis. "As we get more into the broadband Web, our understanding of that will expand by orders of magnitude."

Davis has a year off from teaching to set up the research lab in downtown Berkeley. During that time, Yahoo! will watch to see whether and how to approach other academic researchers for similar ventures.

"We hope there will be many relationships with other universities," says Raghavan. "Our intent isn't to take the viewpoint that the only good research people are our own employees, but to get the viewpoints of people in academia."