• by Meir Zohar , October 8, 2008

There is no doubt that the 2008 Presidential election campaign will be remembered for the use of video. Both campaigns are uploading videos minutes after they were shot, even during the middle of the debates, and running commercials exclusively on video websites like YouTube.

And creating and running videos offers a very cost effective strategy. For no additional cost beyond current staffing and web hosting expenses, the campaigns are generating a lot of buzz for a fraction of the cost of running one 30-second commercial in Cincinnati or Harrisburg.

What's missing from these online videos is the ability to target.

Someone in California or Texas is as likely to see one of these videos as someone in critical election 2008 states like Ohio or Pennsylvania.

But geo-targeting alone isn't enough. Preaching to the converted has never been an effective marketing strategy, and I suspect that most of the people viewing a McCain or Obama video in Ohio or Pennsylvania are probably viewers who have already declared their support for one of the candidates.

The 2004 presidential elections were marked by the addition of a layer of geo-targeting to the campaigns advertising tactics, following the rise of blue state/red state segmentation.

With the advancements in advertising technology and acceptance of targeting, the advertising advancement of the 2008 elections can be behavioral targeting.

Let's think about the process by which voters determine which candidate to support in the ballot box. For the undecided voters the campaigns are currently targeting, it isn't going to be one 30 second commercial that will provide an "a-ha!" moment and then make them decide to vote for Sen. Obama or Sen. McCain. It will more likely be a process of repeated messages relevant to their specific interests that will shift an undecided voter to one camp or the other. You could say that it's a process of appealing to their specific behavior in order to cause a shift towards one of the candidates.

And this is a marketing challenge ideally suited to behavioral targeting.

So what can behavioral targeting offer the presidential election ad campaigns as we head into the home stretch of the election?

1. The ability to segment according to issues and demographics to better target prospective voters: As I noted above, in the 2004 presidential election, geo-targeting played a key tactical role. But in reality, the candidates aren't interested in targeting all of the prospective voters in Pennsylvania or Ohio, just the ones whose vote can be won. And the most effective way is to target according to behavior in conjunction with demographic and geo-targeting. Given the known differences between Obama and McCain voters, the campaigns should know the types of behaviors that are most easily converted into votes in November.

2. Behavioral targeting campaigns work across time in order to address and shift behavioral: Even though there are less than two months until Election Day, there is enough time to run a series of ads to shift behavior. Just look at how the election has changed in the month following the announcement of the respective running mates. A large percentage of undecided voters make their decisions in the last few weeks. This gives each campaign the time to create and run a series of specific issues-oriented ads that will be able to shift perception and convert undecided voter.

3. Within behavioral targeting, an exchanged-based solution offers an opportunity to reach more publishers, and ultimately more users: Given the tight behavioral, demographic and geo-targeting criteria, neither campaign can rely solely on one behavioral targeting vendor. Instead, they should each turn to multiple vendors, or better yet, behavioral targeting Exchanges that work with multiple publishers and ad networks, which will enable the respective campaigns to increase their reach.

The key to an Election Day victory for either candidate is in finding and converting the most undecided voters. And the most effective way to find these needles in the haystack is based on their current behaviors.