How To Measure Attention

Advertising is just storytelling. String together some images and words to prod people to buy products. While the industry can effectively develop messages for TV and traditional media, it is trying to figure out how to tell better stories in digital.

Nearly all people (99%) who see online ads don't click on them. "We are living in a digital world -- we use Uber, Seamless, Open Table -- but the vast amount of things we buy are still in the physical world," said Jonah Goodhart, CEO, Moat, during the Native Ad Forum at Advertising Week 2016. He added that 92% of U.S. retail spend happens in the physical world. There is virtually no relationship that exists between clicks and actual sales, he said.

As it stands, digital has foundational challenges. "It's the equivalent of buying a billboard flipped the wrong way," said Goodhart. "Or it is like watching TV with ads on the side of the screen. "How do you know what part of the screen they noticed?"

advertisement

advertisement

The current set of metrics are unacceptable, he said. Five years ago, the industry gatekeepers -- IAB, 4As, ANA -- defined viewable currency using minimum standards. In order to be considered "viewable," 50% of pixels from display ads need to be continuous for one second; video needs to be up for two continuous seconds.

These guideline are failing advertisers, said Moat. Just 52.9% of desktop ads meet these standards; only 41.4% of mobile ads are viewable. For video, 58.8% are desktop ads are considered viewable and 41.4% of mobile ads. "Human and viewable is the first step, but we must go further," he said. "Attention means noticing something."

Yet this opens another larger debate over how to measure attention. Moat is monitoring most social media channels and has just announced that it will track Snapchat. Still, it remains to be seen if digital metrics can translate as easily as tracking success at physical stores. For instance, you can monitor how many people walk buy the store versus going inside; how much time is spent; active versus passive shoppers, etc. "With those signals I could begin to understand attention," said Goodhart. "The same theory is true in digital," he said. We begin by tracking signals. How much time spent on a page; scroll down; page signals; ad signals; flip to another tab. "The right denominator for a marketer is attention."

1 comment about "How To Measure Attention ".
Check to receive email when comments are posted.
  1. Ed Papazian from Media Dynamics Inc, September 29, 2016 at 11 a.m.

    Attention and/or engagement is a very difficult thing to define, let alone measure. For certain types of exposure situations---say a teen walking to school with a gaggle of gabby pals while using a smartphone to "watch" a short video, an ad might be said to garner some measure of attention if it is noticed and the user's eyes are on it for a few seconds, as that's about all you can expect in such situations. On the other hand a 15-second ad in a long form video that is completely exposed on a user's screen at home on a PC with no one else present, might be judged by some combination of time spent, and other mechanical actions or even better, by an ad recall study---or both types of metrics, human and electronic. The nature of the ad is also to be considered. A silly, "High I'm cool" type of ad lasting three to five seconds is quite a different kettle of fish from a 15- or 30- second story telling presentation by a branding advertiser. In the first case, any sign that the ad might have been noted could be sufficient; in the latter case, much more elaborate and interacting measures, particularly as regards motivating power, are mandated.

    TV advertisers have long been aware of these distinctions and their "copy testing" practices might be studied by the digital people as they offer some interesting insights that merit study. The basic point is simple. If you rely exclusively on electronic indicators and try for a single all-encompassing metric, you will miss many of the vital nuances and the results will not be comparable from one campaign--- or venue or exposure situation--- to the other.

Next story loading loading..