WPP’s GroupM unit this morning said it has partnered with big media companies to create new standards for measuring and verifying the "viewability" of ad on streaming and CTV (connected TV) environments.
The announcement says GroupM and the media companies are working with "standards-setting bodies to create a streamlined measurement framework and set of best practices to ensure ads are only counted when delivered to screens that are on, with people in front of them,” but oddly does not include any explicit reference to the Media Rating Council (MRC), which has served as the ad industry’s self-regulatory body for setting standards and auditing and accrediting measurement services that meet them since the 1960s, following Congressional hearing and a consent decree with the U.S. Department of Justice.
advertisement
advertisement
Odder still, the announcement comes one weekday after the MRC held a webinar explaining its standards-setting process for audience measurement, how it works and why it is integral to the audits and accreditation processes it oversees for the U.S. ad industry.
The only trade body explicitly cited in the announcement is the Interactive Advertising Bureau (IAB), which normally works closely with the MRC on setting industry audience-measurement standards.
"IAB and IAB Tech Lab have been at the forefront of the unique challenges of the CTV space across publishers, platforms, apps and OEMs," IAB CEO David Cohen said in a statement, included in GroupM’s announcement, adding that the Tech Lab plans to release an “open measurement” SDK (software development kit) for CTV in the third quarter of this year, which will “enable the industry to facilitate third-party viewability and verification measurement.
“It will also deliver a 'set off' signal and algorithmic assessment around the likelihood that a person is in the room viewing. We commend GroupM and others who look to quantify industry opportunities such as this one and look forward to continued education, best practice sharing, and technical standards to address these areas. Our Video Leadership Summit happening in just a few short weeks will tackle these issues head-on."
GroupM said the initiative follows a joint study it conducted with audience-measurement firm iSpot that "quantifies inflated CTV ad delivery counts and reveals that, on average, 8-10% of streaming impressions play when the TV is shut off, primarily through ancillary devices."
The GroupM/iSpot study also found:
iSpot has been among the audience-measurement firms vying as "alternate currencies" to Nielsen for the TV and so-called "cross-platform" ratings in this year's upfront advertising marketplace.
iSpot, which has even been independently "certified" as ad market currency by at least one big media company -- NBCUniversal -- currently is the only Nielsen alternative that has any MRC accreditation, though that is only for TV occurrence data, not for actual audience measurement.
GroupM said the initiative includes NBCU, Disney, Fox/Tubi, LG Ads Solutions, Paramount, Vizio and Warner Bros. Discovery.
Historically, "viewabilty" standards in the U.S. ad marketplace have been set by the MRC, which created the original minimum viewable impression, as well as duration-weighted viewable impression standards for both static digital, as well as digital video advertising.
But at least in terms of this viewability standards-setting initiative, the MRC will be taking a backseat.
"Our initial work was to quantify the problem and addressing standards setting is a key next step for us," a GroupM spokesperson tells MediaPost, adding, "We are engaged with the IAB, and working in partnership with the MRC and others to build more transparency and counting rigor into our increasingly complicated media marketplace."
What struck me about these findings, Joe, is that there was virtually no inflation---viewing counted when set off--- for native smart TV apps which accounted for half of all of the "impressions". If that is so and the overall findings were that the average inflation was 8-10% that means that the other half of the impressions, overcounted the "audience" by 16-20%. Gulp!
@Ed Papazian: That's correct. All of the inflated "TV off" impressions occurred in the non-native apps portion accounting for the other half of the medium:
STUDY FINDINGS: The Phase One study found that:
On average, 8-10% of overall streaming impressions were delivered when the TV was shut off.
17% of impressions delivered only through CTV streaming devices such as dongles, gaming consoles, and sticks were delivered when the TV was shut off. Gaming consoles generally had lower rates of continuous play than dongles and sticks.
Native Smart TV Apps—which account for about 50% of all CTV viewing[1]—had virtually no incidence of overcounts across streaming ad delivery.
Depending on the configuration among the three components—TV make and model, streaming device and publisher app—the CTV impression overcount by publisher ranged from 2.5% to 15% across all CTV streaming activity.
Rentrak/Comscore (and other set top box measurement companies) have trouble accounting for TV sets measured but turned off. They did create an algorithm to try to deal with it.
"Algorithm" is 21st Century fancytalk for "guessin'."
Ed, they claimed this doesn't happen when the apps are installed on the TVs themselves. They assumed the apps were off when the TV was off. BUT ...
not necessarily.
in mobile ad fraud, for example, your smartphone screen is OFF, but the phone is still on and connected to the internet
mobile apps have been documented to continuously load ads all night long. for example how many minutes do you use Loudest Alarm Clock app? That app alone is generating more ad impressions than ESPN, and nearly as many impressions as Hulu and Spotify.
Thanks, Augustine. It gets scarrier and scarrier---doesn't it.