DPAC III: View Throughs vs Time Exposure metrics
May 13th, 2009 by Kathryn Koegel
The session was lively with Jon Gibs throwing out the bomb that online display is broken. While we disagree on that point, think we ultimately come to the same conclusions. Yes, a model based on impressions that rewards the publisher for as many ads per page is just bad and does nothing to improve ad performance. Time exposure metrics are a great idea, but only for some types of ads: those designed to NOT drive some kind of action but to engage the user on the page. The reality is, that because online, unlike other media, can lead to so many other destinations, most advertisers using it will have some kind of end goal, even if that goal is not an immediate click. I’d like to see the industry — especially publishers — finally embrace view through (made widely available through comScore), and set the windows appropriately. Time exposure metrics are great, as are interaction metrics that Google discussed — but for certain types of creatives. Kyle Johnson of Compete illustrated this well with two Toyota Venza creatives: one that performed best in terms of delivering consumers through to site, the other that worked better on page. Both had different ROI but showed that as he said, you have to look at more than one measure of success.
The great thing is that we gave attendees some solid examples of how they should increase the sophistication of their analysis.
read the whitepaper at:
www.primaryimpact.com/stateofdisplay.html
Tags: comScore, doubleclick, google, Nielsen, online advertising analytics