Digital media measurement vendor DoubleVerify has spent years ardently stressing the festering problem of connected TV advertising fraud. On one hand, CTV ad fraud has been exacerbated by the pandemic and therefore been good for business, while the other hand shows that the proliferation of CTV viewing has made DoubleVerify’s job of debunking myths a little trickier.
With TV ad dollars shifting post-haste from traditional pay TV to CTV apps, marketers are being lulled into a false sense of security on the pretense that their ad dollars are being targeted at the crème de la crème of targeted viewers. That just isn’t the case for many advertisers.
DoubleVerify’s reporting recently discovered that an unnamed client’s ad campaigns were running in 50 apps across CTV devices, when the client was of the understanding it had signed up for campaigns on just 8 apps. This sounds like getting more bang for your buck, but this additional reach brings viewability concerns because around half of impressions failed to play even 25% of the ad, according to DoubleVerify’s data. The vendor eventually refined the campaigns to bring completion rates up to 70%, but still DoubleVerify’s technology does not solve the fundamental issue of ad fraud at the source.
For that reason, CEO Mark Zagorski took to this year’s virtual CES to outline four challenges to realizing the CTV advertising opportunity. He slammed inconsistent technology and standards across the industry, calling for standardization of the way transactions are tracked and measured between buyers and sellers on CTV. In Zagorski’s eyes, there needs to be one metric to rule them all; one metric upon which an advertiser feels confident in buying across digital platforms.
This dovetails with inconsistent data. “Right now, we have challenges with certain platforms withholding certain types of data that helps with easier measurement,” said Zagorski, without naming any names. “This is very different to linear where a rating point is a rating point, and all publishers or linear TV networks provide as much data as possible to fuel that data (such as set top owners releasing certain amounts of data so metrics can be put together). We need to move beyond data gardens to a world of consistent technology.”
Another challenge that advertisers did not face in linear TV is the ability to block certain types of content, a feature which is still not universally available in the CTV ecosystem, although Zagorski admitted that is changing rapidly.
His final gripe is the issue of multiple tags creating operational inefficiencies. “A single piece of code can work across multiple types of properties. We’ve created an omni-tag so we can capture data on any device. A single tag creates a single view across all media,” added Zagorski.
It’s notable that companies like Nielsen have been challenged to do this, but matching data across set tops, people meters, diary entries and more sources is not something the incumbent measurement firm has seen as worthwhile.
DoubleVerify CPO Jack Smith then took the reins. He drilled down into why CTV is such an attractive prospect to fraudsters – with high CPM and high-demand inventory being the biggest lures.
Dropping in some frightening figures, Smith observed a 161% increase in CTV fraud rate from Q1 2019 to Q1 2020, monitoring over 1,300 fraudulent CTV apps in H1 2020 alone in that time, while detecting over half a million fraudulent devices every day.
An underlying problem is that CTVs are perceived as premium devices, when the reality is that bad apps can still appear on good platforms – and this problem primarily manifests as bot fraud. While bots account for about 20% of ad fraud on desktop and mobile, about 78% of CTV fraud is driven by bots, according to Smith.
As well as doing its part on the education side, DoubleVerify does have some actual proprietary technology for combatting ad fraud. The company has built a dedicated fraud lab with an in-house AI system lead by a crack team of data scientists constantly looking at suspicious activity. This human-machine combination is essential to combatting ad fraud.
The creation of a certification program with some of the top programmatic platforms last year was another milestone for DoubleVerify in stamping out ad fraud and raising awareness. Members include SpotX, The Trade Desk, Xandr, Verizon Media, MediaMath, Amobee, Adephic, and VideoAmp, which must demonstrate to DoubleVerify the ability to prevent fraud and invalidate traffic by applying pre-bid app and device protection.
And clearly it’s working, with DoubleVerify finding 11% less fraud within the certification program compared with buying programmatically through the open market. Smith also warned of seeing 1.2% more ad fraud within non-certified platforms that are direct to publishers.
This program launched almost exactly a year ago, for detecting both ad fraud and invalid traffic on CTVs – using detection algorithms without requiring SDK integration.
Since then, DoubleVerify says it has increased transparency through its reporting, by adding features such as inclusion/exclusion lists and video filtering (including blocking for CTV), while DoubleVerify’s 2021 roadmap is all about building full transparency in CTV environments. This year, it plans to add app-level controls (e.g. star ratings, user reviews), as well as content-level classification and avoidance.
DoubleVerify has already kicked off 2021 by achieving Media Rating Council (MRC) accreditation for display and video rendered ad impression measurement and sophisticated invalid traffic filtration in CTVs, including – of course – ad fraud.
“Our clients want consistent classification and to use the same rule sets used on other channels,” added Zagorski.