Ad Network & Exchange Guide 2009

Will Using Behavioral Data Lead to Smarter Ad Buys?

Aggregators and Exchanges Aim to Create 'Liquid Market' Based on Users' Activities, Not Their Locations -- but Can They Get Past Privacy Concerns?

By Published on .

NEW YORK ( -- Could data about impressions -- rather than impressions themselves -- be the real winner as the online-display industry moves toward an exchange model?

That's the contention of a growing number of so-called data aggregators and exchanges, with companies such as BlueKai, eXelate and Datran on the behavioral side and players such as Lotame, SocialMedia and Media6Degrees working the social-media world.

BlueKai CEO Omar Tawakol
BlueKai CEO Omar Tawakol
They are aiming to create a liquid market for anonymous, cookie-based user behavioral data. For the most part, this is the same sort of data-driven targeting enabled by first-generation behavioral-advertising players such as Tacoda or Revenue Science. The big difference is that rather than bundling impressions and data -- an approach that leads to scaling problems -- second-generation data exchanges traffic only in data, which can be combined with almost any impression in any number of ways.

Data exchanges work like this: Behavioral data are collected on publisher or e-commerce websites (or via offline methods, such as the collection of straight demographic data), then sold to a data exchange, which then resells the data at market value to advertisers, which use the data to buy impressions directly from another publisher via an ad network or ad exchange. The result is that data exchanges can, for instance, collect information from travel site Kayak about users who are booking flights to Orlando, then sell that cookie-based data to a buyer in the exchange. That buyer then uses that data to go to another party to buy and serve ads (say, for a hotel stay) to those users when they are surfing another site. Just as likely, a publisher or network could buy the data in order to improve the ad targeting on the sites they're selling. Social-media targeting works in much the same way, only it captures different types of user activities, such as uploading, sharing or creating user-generated content, actions typical of influencers.

Overall, 65% of marketers said they use or plan to use behavioral targeting, while 74% said they believe such targeting is effective, according to a recent study published by Datran Media. The company surveyed more than 3,000 industry executives from Fortune 1,000 brands and interactive agencies. In a recent report, Forrester Research estimated that one-quarter of online campaigns use some sort of behavioral-targeting data, though key challenges remain in the areas of privacy and data sharing standards, said Forrester analyst Emily Riley.

Value of data
Indeed, while you would think that privacy concerns and the potential impact on existing players might have data exchanges stepping lightly to stay under the radar, that isn't always the case. For instance, BlueKai CEO Omar Tawakol boldly touts his company's potential. "We're rolling a grenade into the room; we are disrupting the ecosystem rather than fitting into it," he said. "We have this stubborn opinion that data can be valued independent of media -- and that in many ways it is more valuable." What's the price for this? Mr. Tawakol told The New York Times that 1,000 impressions bought using Blue Kai data can cost $4 to $15.

Not only is data targeting not necessarily a new phenomenon, data houses have a long history in the offline world, where companies such as Acxiom and Experian have long trafficked in user data. What's new with data exchanges is that they fit in so well with the increasingly automated and efficient Display 2.0 ecosystem of ad exchanges, demand-side agency marketplaces and publisher networks.

EXelate, for instance, collects data on about 110 million unique U.S. users per month, broken down into 16 different targeting segments, acquired through partnership with about three dozen publishers. On the buyer side, it works with about 40 ad networks and agencies as well as with buy-side optimization platforms such as Media Math and X Plus One, said eXelate Chief Revenue Officer Mark Zagorski.

"We're moving rapidly to the separation of audience and [online] location; it's really coming down to buying audience," Mr. Zagorski said. "The user is becoming as important as the context or site where the user is."

While this presents clear opportunities for agencies and their clients to be smarter about online ad buys, the opportunity for publishers is almost as significant, Mr. Zagorski said. He insists third-party data exchanges offer a great opportunity for publishers to monetize not only their page views but, more importantly, their hard-won audiences. Publishers might even be data buyers in some cases. "They can go beyond just selling media now and sell data. It behooves all publishers to have some sort of data strategy," Mr. Zagorski said.

Bundling cookies, impressions
BlueKai, which went live early last year but really started ramping up last fall, focuses on "in-market data" in three categories: retail, auto and travel. It has data on about 100 million unique users in those areas and has integrated with 60% of the top 20 ad networks and agencies; transaction volume on its exchange is doubling about every two months, said Mr. Tawakol. A recent deal with performance ad network Adconion shows the business in action. Adconion buys BlueKai behavioral cookies, bundles them with impressions and sells them to advertisers, combining "media, data and optimization together like never before," said Ben Fox, Adconion's head of product management.

How the online-ad data market plays out remains to be seen. Third-party exchanges hold great appeal, but ad networks could also play a key broker role, while emerging agency "buy-side" networks and exchanges will also be capturing valuable user data which will grow even more valuable over time.

The wild card in all this is the concern over privacy. The FCC and FTC have indicated they will watch behavioral advertising closely, and the concept of an exchange buying and selling data about users could certainly draw unwanted attention. But with the publishing and ad industries hurt deeply by the downturn, the counterarguments are also gaining ground: Why shouldn't publishers get paid for their audiences and content? Why wouldn't advertisers want to target the customers they truly want to reach? Data exchanges could pay a key role in answering those questions.

Most Popular