Behavioral Focus: Hanging On the Telephone

It's all about who you know

As recently as five years ago, the basic Internet monetization model was one that even the oldest of the old school, Time Inc. magazine empire founder Henry Luce, would have recognized. Vertical content (health, auto, travel, etc.) is shaped by professional editors and, therefore, easily categorized by the taxonomy engines that informed the first generation of online ad targeting solutions from pioneers like DoubleClick, Google and Tacoda.

But the past few years have shown that the pros aren't the be-all, end-all anymore, least of all online. By some estimates, nearly 70 percent of consumer time online is spent engaging not with professionally published content but with material created by fellow consumers, including blogs, social network pages and photo-sharing sites. What's even more striking is that up to 60 percent of all traffic within these environments takes the form of communication between members. The big user-generated content platforms like social networks, photo-sharing sites and even blogs facilitate, at their root, communication.

So what does history tell us about the successful monetization of social network content both inside and outside of the original channel? Naturally, the data comes from one of the owners of the original social network: a telephone company. In fascinating research conducted in 2004, Chris Volinsky, the director of statistical research at AT&T Labs Research, undertook a study with Foster Provost, a New York University business professor, and Shawndra Hill, then an NYU graduate student and today a professor at the Wharton School. What these researchers discovered was remarkable: Any person in contact with an existing customer of a firm is three to five times more likely to respond to a message from the firm. Birds of a feather do indeed flock (and buy) together.

Even more significant, the researchers recorded these results in a direct-mail channel that involved neither an explicit nor an implied endorsement by one consumer to another - à la the troubled Facebook Beacon. In essence, the researchers found that by analyzing which customers communicated with each other, using inbound/outbound pairs on the telephone grid, they could identify "network neighbors" in the telephone social graph.

If they found one network neighbor to have responded to a particular direct mail offer, then sending the same offer to his network neighbors resulted in a three- to fivefold lift above any targeting technique not informed by this network-neighbor data. The researchers explained the results this way: "Social theory tells us that people who communicate with each other are more likely to be similar to each other, a concept called homophily .... Linked consumers probably are like-minded, and like-minded consumers tend to buy the same products."

Just six months ago, people dismissed the marketing potential inherent in understanding the composition of nodes in social networks. But I'm confident that such social data will soon inform the online media strategy of sophisticated buyers in a significant manner, and they'll even challenge the dominance of demographic data as a targeting method. When it comes to targeting advertising, the AT&T study indicates that who you know is more important than where you live.

How does one re-create the AT&T experience in the context of high-volume online ad-serving using social network data? How can marketers effectively leverage this data while maintaining the highest standards of consumer privacy?

The answers are not yet clear. But the directional data inherent in the AT&T research is remarkable.

Over the next 12 months, sophisticated media buyers will increasingly ask us if we know our neighbors.

David Honig is a cofounder of Media6Degrees, a New York-based firm specializing in the use of social network data for the targeting of online advertisements. (

Next story loading loading..