Targeting By Interaction
Behavioral targeting has largely been geared to increasing the odds that ads will be served and seen by the right prospects. So far, so good. Unfortunately, as Ian Swanson, CEO of Sometrics, explains below, too often behavioral targeters think their job ends when an ad is served. With social networks in particular, he says, that's precisely where the real work begins.
Behavioral Insider: How did Sometrics get started?
Ian Swanson: My experience with social networking dates back to being involved with Userplane, which grew to be one of the largest social media properties and was acquired by AOL. We had tons of social ad inventory, but lacked an effective understanding of the audience engaging with the advertisements. The challenge then, which has been one all the social networks have dealt with ever since, has been how to effectively understand and monetize the social Web.
The first step for Sometrics came last May, when Facebook opened up its applications development platform. It was all one big sandbox. One of our partners had taken part in the Human Genome project, and so what he explained was that we could take a similar approach. In bio-informatics you take isolated data and work back to find the relevant connections between data points.
BI: What motivated the introduction of your analytics apps, and how can developers deploy it? How do you distinguish what you're doing from the platforms of traditional BT technology platforms like Tacoda or Touch Clarity?
Swanson: What we did was begin working with developers on what we called social analytics. Most of these developers had little or nothing in the way of analytics except very rudimentary traffic information of the kind Google gives them. What they didn't have was the kind of traditional high-end deep analytics of a large Web site -- the kind, say, an Omniture offers -- applied to tracking and generating social media specific metrics.
So we addressed the challenge of providing the type of data application developers would really need to monetize. We've worked with about a thousand developers over the past year. Specifically, we help them see who the audience is that's using their applications and how they're using them, providing access to data about installs and uninstalls, age, gender, geography, interests -- and beyond that, social specific data like how many friends they have.
BI: You also do ad-serving?
Swanson: The next step we've taken is to expand the platform to include ad-serving. So what that means, firstly, is that social Web developers can target ads based on the analytics we provide, by demographics, by behavior, by geography. But beyond that it means that they can learn how their audience engages with the ads that are served. That's data advertisers are starved for, especially in the social media space. Our tools are geared to telling who interacted with ads in a variety of ways. For instance, we could tell a Nike that it's primarily 18- to 24-year-olds who are not only looking at their ads but adding them to their profiles or sharing them. That's highly relevant information that isn't just captured in click data. For brands looking to build engagement and affinity, it's deeply important to know things like who's not only clicking but saving and sharing data.
BI: What kind of feedback have you gotten so far from the agencies and advertisers?
Swanson: There's an educational curve involved anytime you have to change methodology, but we've seen an encouraging take-up by brands looking to push for change in how they approach advertising in a space where click-through has so far been low but engagement high.
BI: What are some deployments on the immediate horizon?
Swanson: One thing we've introduced is a social ad click map, which ties click and engagement data back to demographics. So we can tell an advertiser who's trying to target young males that a certain creative gained clicks and high engagement metrics with males 27 [years old], who live in Southern California, like the Lakers and Green Day. This is a new type of ad analytics made possible by social profiles and not available to agencies and brands till now.
In addition to using that information to targeting ads on a social network, they can find out who's truly interacting with their ads. That's information that makes a social media publisher or developer's inventory more valuable, yes, but it's also data that can be leveraged all across the Web. Brands are realizing they can get marketing study value that would cost tens of thousands by buying $1 CPM social ads through Sometrics.