Eyeblaster GM Yoav Arnstein walked iMedia Brand Summit attendees through better ways to understand consumer engagement in a world of ads that move.

Is the Emperor Naked? That was the question posed by Yoav Arnstein, Vice President/General Manager of Eyeblaster, Inc. earlier this week at the iMedia Brand Summit in Bonita Springs, Florida. It was his way of asking if media buyers are using the right set of metrics to measure the success of rich media units. He cited a wide variety of rich media measurements used by the online industry, such as, “Interaction Rate is the percentage of audience that interacts with an ad…” (Kagan); “…The percentage of the audience that engages with the ad unit,” (DoubleClick); and “This rate is determined by taking the total number of unique interactions per impression divided by total ad impressions” (PointRoll).

But Arnstein noted that none of them truly measure the value of user interaction with brands. There are just too many variables.

For example, Arnstein said an ad unit’s call-to-action options can affect reporting results showing how an expandable banner with a questionnaire (forcing greater interaction) outperformed a floating ad with one clickthrough option by more than double.

Arnstein said that many reporting methods do not distinguish between positive and negative user interactions such as closing the unit, muting the sound or stopping the unit before it has fully played (such as video). Moreover, some rich media units may launch automatically as soon as the user comes to a page or mouses over an ad, but they are never actually seen by the user. These false/positives are often included in interaction rates.

It is more important, Arnstein said, to measure how and why the user interacts with the ad-- especially when they engage with it more than once (such as users coming back again to watch or re-watch a video, later to download a screen saver and a third time to enter a contest). These multiple exposes mean that users have engaged for a longer time with the brand. Although users who engage for a longer period of time with the brand message are only counted as one unique visitor, they take intended actions at a far higher rate than casual visitors, and thus should be weighted as more important in measuring the success of the unit.

Arnstein argues that it is important to strip out the negative interactions and visitors who unintentionally auto launch a unit (that mouse-over again) and to understand multiple exposures before judging the true impact of a rich media ad.

What at first may appear to be a mediocre performance can be discovered to be a highly-effective unit when factoring in if the user took an intended action.

Arnstein says that where a rich media ad runs can impact the user interaction rate, further saying that visitor-to-user-generated-content sites tend to engage longer with rich media units than. say, on an entertainment site.

Since different sites will provide differently response rates to different kinds of rich media, it is important that buyers don’t benchmark ad success by a limited site schedule.

When asked if rich media should move to a PPC model since there are so many variables that can impact the response rate, Arnstein said with a smile, that he would be happy to do that, “when I can also control the creative.”

George H. Simpson has been an expert in advertising, marketing and Internet trade media relations, for nearly 35 years. Read full bio.