Are Those Really Behaviors You're Targeting? Think Again! (Part II)

by Robert Tas, Nov 22, 2004, 12:00 AM

• Comment
• Recommend (2)
• Tweet
• 

Last week, Robert Tas discussed the definition and scale of behavioral targeting. The continuance of that discussion is below.

Performance

Performance is the "you'll know it when you see it" end of behavioral targeting. Let's face it, none of us know what the right performance is, and I would suggest performance be looked at from a few perspectives.

One certainly should be the click rate from the targeted ad. The click rate for behaviorally targeted ads will likely be higher than an untargeted ad. (How's that for stepping out on a limb?) Data from our initial campaigns show a 25 percent favorable lift from untargeted campaigns, but we are early in the cycle of analyzing the data.

Performance in behavioral advertising can also be viewed from a reach and frequency standpoint. The advertiser should be asking, "Am I exposing my message to a large enough audience with enough frequency to drive a measurable change in brand perception?"

The final view of performance is the most simple: do the ads targeted to the behavioral segment drive more sales? In many cases, an increase in click rate does not correlate to an increase in sales. The basic advertising rules of persuasion and purchase must be adhered to.

Integration

Judging how the behavioral target is integrated into the overall ad buy for an advertiser or the enterprise wide sales strategy for a publisher is critical for setting up measurable goals for the success of behavioral targeting projects.

If behavioral campaigns are treated as one-offs or exceptions, the full benefits of an integrated sales or media plan can never be realized. It is key for publishers to decide which areas of high demand drive data collection and how this can increase low demand inventory sell through.

Segmentation

Segmentation begins with your audience and your advertisers. Decide what to measure, how to measure it, and then start measuring. It is a mix of content interactions to establish the conventions of the behavior measured and number of times a member visits before being allowed to join a segment. We use a simple formula: the frequency of an interaction over time equals a behavioral segment. The exact value for each variable may change depending on the segment being built but the formula remains the same.

Deciding which interaction you want to track is simple enough, but the secret sauce is in how high a bar is set for segment membership. Make the frequency count too high and the segment is too small to matter. Define a lengthy time frame and the segment becomes homogenized causing low response due to irrelevant ads being served.

When used as a framework for comparison, the five elements of an effective behavioral targeting solution will help publishers and advertisers set proper expectations and realize success together. Behavioral targeting is all about delivering the right ad to the right person, so they can buy something. With the right perspective behavioral targeting won't get lost in the hype and will become a standard tool for marketers.

• Comment
• Recommend (2)
• Tweet
• 

Be the first to comment on "Are Those Really Behaviors You're Targeting? Think Again! (Part II)"

Leave a Comment