Prabhakar Raghavan, Research Chief, Yahoo

Prabhakar RaghavanPrabhakar Raghavan joined Yahoo as head of research on July 28, with a mission to maintain the company’s lead in consumer services while eating away at Google’s lead in the search sector. He has the chops to do it.

Raghavan came to Yahoo from Verity , where, as CTO, he led the development of its business portal software, including developing automatic classification algorithms and the concept of the social network as the foundation for the business portal infrastructure.

Prior to that, Raghavan headed the Computer Science Principles department at IBM’s Almaden Research Center; he led IBM’s CLEVER project on Web search and mining. CLEVER was an early Web search tool that analyzed links to determine the most relevant results, the method that made Google a household word. He holds several patents on link analysis for Web search.

Q: How do you think Yahoo stacks up today against its rivals in terms of services and technology?

Exceptionally well in terms of the engineering backend. Some of the surface sizzle could certainly work a little harder. We’ve been putting a lot of effort into the backend, hiring Larry Tessler, a user interface guru from Xerox PARC a couple of months ago. Our intent was to dramatically emphasize the whole user interface side, which has always been pretty good.

We’ll make substantial investments in the area of human/computer interaction, which is assuming special significance now that we’re no longer just talking about the browser and Windows interfaces. We want a richness of experience that persists across the living room and mobile devices.

Q: According to the corporate Web site, Yahoo has around 30 researchers? Will you try to increase that?

You can expect to see a significant growth. It’s not like there’s a budget we’re shooting for, but a particular quality. Once you set the bar high enough, it’s not a question of budget, but can you hope to hire people at that level of quality?

Our view is that research isn’t only done in my organization, but in engineering units, as well. In addition, we have expanded our network with academic institutions. The Berkeley Laboratory is the first of its kind, and we expect to roll out more in campuses across the world. Also, we’ll announce some very recent hires in the next few weeks.

Q: When it comes to competing for talent with Google, does Yahoo face a cachet gap?

In our particular target segment of the population, it’s not obvious. I’m looking for fairly hard-boiled scientists who can look past cachet and cute IQ tests. As we hired people who looked at both opportunities and made the choice to come here, the appeal is the ability to do world-class science, do it openly and to publish, which is a huge deal for the eminent scientists we’re trying to attract.

They don’t want to be told that once hired, they have to give up academic work and focus on products. That’s a huge deal, and many have told me it’s a big attraction of coming here. As we add more “names” to our roster, it will add to the rich-get-richer phenomenon.

Q: How much focus will you place on search as opposed to other services?

It’s a little hard to quantify. Obviously search is a big part of our business right now, but we’d rather not put all our eggs in one basket.

One of the first things I had to do was figure out the scientific disciplines we’d invest in. Obviously, one is around search and information retrieval. That’s a key part of our business now and in the foreseeable future.

Then, there’s the whole area of social media. What do you do when you have essentially a billion people creating content and consuming content? How do you aggregate this content, such as Web pages, video blogs and podcasts in a rich way and act as an intermediary, and hopefully monetizing the process?

We have a specific notion of social media here: Instead of thinking of market segments at a fairly gross level, breaking the population into 10 or 15 buckets, the world we’re interested in is very different. You can break up the population into a billion different categories and ensure each one gets a different experience.

There are group dynamics of creating content and showing it to others. Understanding how communities of interest ebb and flow is to understand the economics of these communities. How do we monetize this? If you have a chain of people creating content and annotating it, and it even makes money, how do you apportion the money? This is the kind of scientific endeavor that has business consequences.

The final scientific discipline we care about deeply is large-scale, distributed computing systems. We’re no longer talking about 20 machines in a lab, we’re talking about a million machines spread around the planet to create the fabric of computing.

Q: IBM is known for its huge patent portfolio. Will you try to increase the patent stream for Yahoo?

We need to do that. It’s one of our strategic goals and something the company as a whole has to focus on. It’s equally the job of all technical leadership in the company, and we’re all doing that constantly.

Q: What part of the job are you most looking forward to?

Yesterday, the world in was sitting at a browser and retroactively looking for information. Going forward, there will be multiple touch points, including car information systems, mobile browsers, and the living room. Not as much information will come to you in response to search; a lot will come to you proactively.

We need to facilitate it coming to you proactively with the right relevance and volume, so that it engages you without overwhelming you. At the backend, the volume of the world’s information is growing exponentially, while your ability to consume it is fixed. That means we at Yahoo have to get exponentially better at funneling down to you the information you want. And that’s the opportunity I’m challenged with.

News Around the Web