DoubleVerify — the advertising measurement company that identified last year that one in four CTV environments and apps continue to play ads when TVs are turned off — is now offering what it describes as the first scalable measurement solution for the problem.
Following DV’s report, a study by GroupM and iSpot.TV — which have partnered to develop their own TV off solution (more on that below) — found that the problem results in streaming impressions being overcounted by an average 8% to 10%. That resulted in about $1 billion in wasted advertising spend last year, Stephens analyst Nicholas Zanger has estimated.
Fragmentation across devices, time and locations and lack of technical standardization has hampered CTV ad viewability measurement.
Although a "TV off" signal is sent across all CTV devices, not all CTV applications heed it, so they continue to play content and deliver what are incorrectly recorded as paid ad impressions.
I don’t pretend to be a tech expert, but I think some of the basics are important here, so please bear with me as I attempt to summarize how DV’s solution works, with help from DV, and the broader CTV viewability measurement scenario.
IAB’s VAST, the industry-standard digital video ad serving template for structuring ad tags that serve ads to video players, was supplemented by VPAID (Video Player-Ad Interface Definition) as the standard used to measure engagement for digital video and enable programmatic delivery and delivery of ads with complex interactions. But VPAID is not supported for CTV. (And is in any case being replaced by new standards including an open measurement software development kit/SDK and open measurement interface definition and a Secure Interactive Media Interface Definition, none of which I’m about to get into here.)
DV’s Fully On-Screen Completion (FOS) tool was launched a few years ago to be a technical proxy for viewability measurement in CTV, and was independently evaluated and certified by the Media Rating Council (MRC).
DV’s enhanced version of FOS measures how an ad loads — whether an ad impression meets the IAB’s industry standards for what percentage of an ad’s pixels must be in view, and for how long, for it to qualify as having been viewable. It also allows advertisers to track viewability directly in CTV — an industry first, according to the company.
The enhancements come from combining FOS measurement with data on the total length of the video ad. DV can now directly evaluate if a CTV ad is “fully on-screen” (100% of pixels rendered) for two seconds or more. That goes a step beyond what is traditionally considered viewable (50% of pixels), and enables a direct video viewability metric and apples-to-apples comparisons across channels and video types, including the up-to-now siloed CTV, says DV.
DV’s technology also factors in TV off signals to understand if ads are potentially running while TVs are off. DV calls environments, or
apps on specific devices, that do not respond to the TV off signal “power unaware,” and environments that do respond “power aware.”
On the “real person” front, another DV ad-fraud technology can estimate non-human and bot-driven ad renderings or ad interactions. Combined with the viewability tech, DV says it can tell if an ad is viewable and if bots are forcing a render. If it’s viewable and deemed human traffic/not a bot, then the ad is considered “viewable and by a real person,” or what DV calls an “Authentic Ad.”
DV says that “a number of leading brands” are already planning to use its new viewability solution, but declined to elaborate.
The GroupM/iSpot.TV study mentioned above found that CTV impression overcounts due to TV off ranged from 2.5% to 15%, depending on the mix of TV make and model, streaming device and publisher app.
Nearly no overcounting was found when the operating system embedded in a smart TV was used for measurement.
Samsung, Vizio and LG operating systems are accessible only through TV integration —giving them a measurement advantage over CTV operating systems available from Roku, Amazon, Google and Apple, which are accessible through sticks or dongles, Zanger has said.
Based on the iSpot.TV study, GroupM last June announced that it and Walt Disney, Fox/Tubi, Paramount, NBCUniversal, Warner Bros. Discovery and Vizion and LG Ads Solutions would work with agencies and advertisers and standard-setting bodies to develop a new industry standard on a new CTV viewability standard.
At the same time, GroupM and iSpot announced the creation of iSpotCTV Verification, a system that promises to be able to measure continuous play at the campaign level and record impressions only for ads on screens that are turned on and being watched by people. The offering is built on iSpot’s Unified Measurement product and will reportedly employ proprietary technology for verifying ad and content play directly on TVs.
GroupM was to have exclusive access to the product during its testing and development, after which the software would be launched in the marketplace.
One source says its glass-level verification could allow for measurement across publishers and devices, which could challenge DV’s solution, which verifies certified apps on platforms supporting VAST tags.
But competition on this front may not be an issue for a while yet for DV, which recently gathered more buzz when it was tapped, along with Nielsen, as a measurement partner for Netflix's new ad-supported tier.
Asked about the status of the new standards and iSpotCTV Verification, an iSpot.TV spokesperson would only state that “Since revealing the results of the Phase 1 study and beta product, iSpot has continued to collaborate with platforms and agencies to refine and operationalize the CTV Verification product. It is currently in private beta with multiple agencies.”
For more background on the potential complications posed by multiple groups being involved in standard-setting, take a look at an item posted by MediaPost's Joe Mandese last June, when GroupM and iSpot made their announcements about CTV viewability.