Can Social One-Up Behavioral Targeting?

We’re at a run rate of $35 million. Most people BS about that or talk about being big but don’t name a number. We offer performance at scale using social targeting. We’ve just scratched the surface. We take a customer base of a client and we find where those browsers are clustering on the Web that’s disproportionate to a random audience. That method is comparable to Pagerank for search. It’s a breakthrough. It says stop relying on the old method mandated by tradtional media. In search it was library science, which AltaVista was based on. We think what we’ve done is comparable for finding audience for brands. Throw out demographic and contextual and the old ways of finding who you think your audience is and start fresh with something empirical and driven by the digital medium. We can use the power of the digital medium and data science to find who those prospects are, who we expect will be your future customers.

It matters. More data is better. We get mapping relationships from publishing partners, but it is confidential. But what we do with it is the key. I view the data we contract for as low-test fuel. The engine we’ve built is a highly sophisticated engine that works on it. It’s commodity data, but the more of it there is, the better. It’s what we do with the data that distinguishes us. The data science that figures out how to score a browser is really complex and highly empirical. We started with theories then we developed methods that drove conversions.

[Behavioral targeting] still has a place. It got its big boost five years ago from the auto industry. It was finding intenders. They ended up responding better than non-intenders. The problem is you’re trying to sell Mercedes and your intenders end up being people interested in Chevy Camaros. Our view is it doesn’t scale very well and if you push it to scale it doesn’t perform. We’ll take those intenders and find the people who share brand affinity because we’ve been able to identify a proxy for their interest based on where they cluster on the Web. It’s more powerful and scalable. We always say put it to the test and see how we do.

They haven’t done any of this work. I honestly wish they would. They have a lot of data that they won’t share and shouldn’t. They have a risk area where they know too much. We don’t. They know so much more. It could get in front of the privacy train. We know so little we’re happy to tell anyone who cares what we know. They haven’t started down this path.

I do think the focus on it is important. Ultimately we all want consumers to feel comfortable with what’s happening. The industry needs to get religion in terms of notice, transparency and opt out. The adoption needs to be better. We went in early and put notice in all our ads. We’re a leader in that arena. We think it’s important. The Wall Street Journal has provided a service by shining a light on it. I tend to think they’ve gone overboard at times. But I think reason will prevail. We’re trying to deliver more relevant advertising to consumers on behalf of marketers, Everyone will benefit if we do it well. Politically the politicians will end up judging it as interesting but complicated and an area where they need to be careful of not stopping innovation.