TARGETING
Published: July 31, 2008
Why your BT campaign doesn't deliver
 

Marketers sometimes lose sight of the role BT plays in their integrated plans. Underscore Marketing's president examines why this happens and how to stay on track.

The best behavioral targeting campaigns start with a comprehensive analysis of a target audience.

That's not the most profound statement in the world. Any ad campaign with a media plan worth its salt conducts a deep dive into the intricacies of the target audience. The difference I want to highlight is how the target information is used when planning a traditional media campaign versus an online behavioral targeting campaign.

In many cases, the details of a target audience analysis that goes beyond the basic age and sex demographics of the buying target are just nuances that help with picking media vehicles. Knowing that your target tends to like jogging as a leisure activity might give "Runner’s World" an edge on your media plan over "Working Mother," or it might lead you to tweak your TV daypart mix to accommodate morning joggers.

The point is that on the traditional side of advertising, this type of target information tends to serve as fine tuning on the sights of a bazooka.

When we talk about online behavioral targeting, though, this information is used to aim a laser.

I've written in this space before about how to look for behavioral predictors or web behaviors that indicate various life stages, but these are targeting methods that are somewhat less precise than directly identifying and targeting desirable web behavior. For instance, I can identify what I'm almost sure are in-market car buyers in Akron if I look for people who input Akron ZIP codes when they configure and price-out a car online. Looking for people who buy homes in Akron is another way of getting at them, but it's less direct.

When we directly observe the desired web behavior that causes someone to fall into the target, we're aiming a laser. That's important to remember when we think about the differences between our targeting campaign and our television campaign. When we think about television in Akron for a car dealership, our train of thought tends to be something like this: "Anyone in Akron over 18 can buy a car, so if we cover the Akron metro with X GRPs per week, we'll keep our dealership top of mind. Let's try to be in daytime, early fringe and a little bit of prime time."

There's little thought given to identifying prospects or just how to most efficiently cover a market.

But when we think about behavioral targeting, we take more of a bottom-up approach. We first think about what kind of desirable behaviors we want to go after, and then we think about how many people in Akron fit the criteria that define those behaviors.

Here's a consequence of thinking about these two media approaches the way we do. In the first example, the currency that governs the media buy is the 18+ population of Akron, which is a much bigger number than the number of people from Akron going to car sites and configuring the car they'd like to buy.

You would be surprised by how often this slips advertisers' minds. They look at how much media weight is delivered by a less-focused traditional media plan, compare it directly with the exceptionally well-planned behavioral targeting campaign, and then they wonder why the targeting campaign seems to have trouble delivering. It's more of a currency problem than a scale problem, if you ask me, but a problem nonetheless.

It's a problem that's best addressed by sizing the market appropriately, which is an important exercise that should be baked in to any target audience analysis. Not only do marketers have to gauge their market potential, they have to think about how many prospects are in market at any given time. This is sometimes skipped over if a television campaign is a foregone conclusion. But marketers can't cut corners here if they want to effectively make use of behavioral targeting.

Tom Hespos is the president of Underscore Marketing and blogs at Hespos.com.