Yahoo Vet Aims to Make Search More Relevant

Greg Coleman Discusses Google, and Why Microhoo Might Be Fun

By Published on .

Like many veterans of portal life, Greg Coleman is making the transition to a start-up -- looking for something "fun, potentially big, and where there'd be a lot of building." After seven years leading worldwide sales at Yahoo, he's now managing a staff of 27 as CEO of NetSeer, a technology company that aims to improve the relevance of search advertising.
Greg Coleman
Greg Coleman

He recently talked with Ad Age about the promise of his new company, why he thinks the Microsoft-Yahoo merger could be good and why unseating Google isn't just "building a better mousetrap."

Abbey Klaassen: What exactly is NetSeer?

Greg Coleman: It was started by some PH.D.s out of UCLA who, three-and-a-half years ago, were trying to discover the relationship of communities on the internet and how they came together. A year ago they had a Eureka moment and realized ... they've been able to map the internet differently than Yahoo and Google. They have indexed billions of pages online and have the ability to create a knowledge bank that understands the words and phrases on internet pages rather than individual keywords. We can find other ways to look at what somebody is looking for based on what's been written on pages billions of times across the internet.

Ms. Klaassen: What's the significance of that in terms of search marketing?

Mr. Coleman: If you look at Yahoo and Google, more than 50% of the search queries go unmonetized. If you look at a phrase like "North Pole" there are no ads because Yahoo and Google do it -- and you notice I still put Yahoo first -- by association of that individual keyword. We've done a bunch of tests that indicate this technology has a number of properties, but one of which could be to monetize pieces of internet that haven't been monetized before.

Ms. Klaassen: Such as the searches for keywords that don't turn up ads.

Mr. Coleman: Yes. Another example is a search on "pigeon's blood." What people don't know is that also refers to a gem, a rare ruby. And it's also a fish. The other search engines wouldn't understand how to derive that. But we have those classifications. ... If the technology is as scalable, we could create a great business not just in doing work with the portals but the entire arbitrage world -- the parking-domain companies, anybody looking for a lift based on getting additional ideas for the particular query that they have. ... In a world where everybody's constantly improving monetization and efficiency and relevancy, we've found something that's different.

Ms. Klaassen: Google has talked publicly about how it is showing fewer but more relevant ads. Does what you're doing contradict that philosophy?

Mr. Coleman: It's complementary. We've not had meetings with the G-word yet but their points are right. It comes down to relevancy. It's not that they don't want to show ads, it's finding ads that are relevant. Particularly as you go down the tail. ... If you can monetize an innocuous term like "pigeon's blood" or "North Pole" or "Cuban cigar" better than you can right now, that's providing the user with something more intuitive than what they're currently getting. So I agree that fewer is better but better is better.

Ms. Klaassen: How do you see consolidation affecting the paid-search market?

Mr. Coleman: It'll bring innovation and innovation should bring lots of opportunities for both the publishers and the users.

Ms. Klaassen: So it sounds like you think a combination of Microsoft-Yahoo is a good thing?

Mr. Coleman: Let's see, how should I answer this? I think the combination of Microsoft and Yahoo could be a very good thing, if they make sure they have the right people in place and a relentless focus on integrating the two companies. It's an "if." It could be great if they had the right senior management on it, with a team mentality to focus the best and the brightest. Under those circumstances I think it would be great. If it fell short, I think it could be a problem.

Ms. Klaassen: You've had some experience with integration—namely pulling together Yahoo's display and search sales teams. Any advice?

Mr. Coleman: You need the right leadership in there to get it done. Secondly, you have to make tough decisions early. Third, if you have people who are going to block an integration, they've got to go. Then you need the DNA to really take over and permeate and to keep it moving. Let's say those companies did get together -- you'd have to have a relentless focus on the customer. ... We did the search and display integration not to save money but to create a better customer experience.

Ms. Klaassen: Do you see any search company unseating Google?

Mr. Coleman: It depends on what happens with Yahoo and Microsoft. But it's different today. When I first got to Yahoo in early spring 2001, Yahoo had made the decision to outsource search to Google. In today's world, that never should have happened, and for any company to open the door like Yahoo did for Google so widely is pretty difficult. The name of the game is distribution. It's to be able to create a better mousetrap but also to create distribution.

Ms. Klaassen: As you see all the web giants building data-driven platforms, do you think that's a good or bad thing for the brand-marketing side of online advertising?

Mr. Coleman: I watch the moves of different companies and players and think you need both the relationships with the advertising community as well as the technology to pull off some of the more sophisticated important things. ... Especially when you're trying to lead innovation. Leading innovation inside one's own conference room is generally doomed to failure. You have to have an exchange with the client. Not only are relationships not going away, but they're more important than ever because people are moving away from it. When you sit down with CMO or VP-sales you need to understand what's important to them, what drives their strategy, and then say "Do I have some stuff in my arsenal that can help them?"
Most Popular
In this article: