“It just feels like a bit of bait and switch,” the buyer said.
The buyer backlash comes at a time when Google, a long-dominant force in the advertising industry, is under particular stress. Google laid off 12,000 employees earlier this year after tepid earnings, and the company faces antitrust suits from multiple jurisdictions. Moreover, the company had already rankled buyers with its Performance Max buying tool, which places ads on Google’s vast swath of media properties via AI, causing concern to some who say it lacks transparency.
Google did not respond to a request for comment by press time, though the company did publish a blog post shortly after the Adalytics report came out, calling its claims “extremely inaccurate.” The post said that in the majority of video ad campaigns served on YouTube, GVP inventory enhances campaign performance and is more than 90% viewable, among other rebuttals.
The performance dilemma
But, removing spend from GVP has consequences, and buyers divesting from the third parties YouTube sends their ads to are worried performance will suffer.
“We turn off GVP, so performance is going to tank,” said one of the media buyer sources.
This is partly because GVP inventory cannot be opted out of for certain performance-centric campaign types, like video action campaigns.
A fourth media buyer said that their agency has been monitoring Video Action Campaign buys since YouTube made GVP inventory the default more than a year ago. Throughout, the buyer said GVP inventory has never made up more than single-digit percentages of campaign inventory and has performed comparably or better than YouTube on return on ad spend and cost per acquisition. This buyer’s experience contrasts with the findings of the Adalytics report, where some brand YouTube campaigns had as much as 84% of inventory via GVP.
“GVP for business goals works pretty well,” said the media buyer, who is not altering their YouTube strategy following the Adalytics report.
Other sources said third parties’ supposedly good performance on lower-funnel metrics is misleading. The open internet is filled with so-called made-for-advertising sites, research has found, which are designed to game the kind of metrics campaigns optimize for but actually have very few engaged viewers.
“In theory, a lot of YouTube works, however, this is an area that makes me question whether a media mix model can be gamed by buying tons of impressions for very low CPMs,” the media buyer source said, noting that YouTube often comes up in the top three in a brand’s mixed media model based on return on investment. “By doing so, you inherently make it a math equation that drives a higher level of return, yet no one ever heard the ad that was delivered below the fold and sound off.”