ChoiceStream

I was starting a blog post in response to John Squire’s recent MediaPost article about attribution tracking and Coremetrics’ new Impression Attribution product when I ran across this survey data from an iProspect study:

What I like about this study is how it takes a holistic view to measuring display ad attribution. It doesn’t just measure clicks and view throughs; it also measures the searches an ad generates on search engines, social media and message boards. Unfortunately, though, this type of comprehensive attribution tracking is not the norm for online display advertising.

One of the hardest things about driving a new solution in the display advertising space is the varying degree of expertise and sophistication around attribution tracking. It amazes me how an industry that has tried and true models around attribution modeling for TV and outdoor advertising does not have even a basic set of agreed upon best practices for measuring the effects of display advertising.

We recently had some disagreement on my ad:tech panel around the value of display advertising as a customer acquisition tool. I realized that the argument came down to how the different panelists were attributing value. Some were using simplistic ‘last click’ based tracking which is hard for networks to perform against and which, in turn, pushes networks to ask for attribution around view through conversions. And we’ve all seen cases where too much focus on view through conversions has led to some shady practices, as David Kopp and I discussed in our recent Retargeting Renaissance webinar.

In far too few cases are advertisers working with their networks to look at the broader effects of their display advertising, beyond clicks and view throughs. Even fewer are running controlled A/B testing that measures display advertising’s impact on search the way the iProspect study does.

And that’s really too bad, because there is a proven, measurable link between display ads and search behavior as illustrated not just by the iProspect research, but also by a robust set of studies done by Yahoo! a couple of years ago. That research showed that users exposed to display advertising demonstrated:

⇒ 26% increase in trademark search
⇒ 60% lift in clickthrough rate on search results
⇒ 35% lift in conversion rate after clicking on search results

It’s true that these metrics can be hard to measure, but they’re not impossible. If we, as an industry, can measure the revenue impact of roadside billboards and 30-second TV commercials, we can surely figure out how to accurately measure online display advertising attribution.

I’d love suggestions as to best practice articles and studies around attribution tracking. I am also in the process of lining up a webinar on the topic, pulling some industry expects into the fold with very practical, prescriptive and actionable advice for online marketers. Stay tuned.