YouTube postpones co-viewing measurement plan

“It was supposed to roll out first quarter [of 2024], and then all of a sudden now it’s rolling out fourth quarter, and there’s no real answer except they’re not ready, which I get. If you’re not ready, you’re not ready,” said one agency executive.

“What is their methodology? What is their panel? We have a lot more visibility into the methodology and framework of other measurement companies, not Google,” said a second agency executive.

“We introduced co-viewing to give advertisers a more comprehensive picture of their audience on YouTube. We encourage advertisers to test and compare campaign performance with and without co-viewing, which is why we’re extending the experiment phase ahead of billing. In line with the way we measure and bill today, advertisers can choose between YouTube or a third party, like Nielsen,” said Kate Alessi, managing director of global product solutions at YouTube and Google, in an emailed statement.

Ad buyers have been opposing YouTube’s co-viewing measurement plans since the summer when the platform informed advertisers and agencies that it would transact against self-reported co-viewing measurement figures. YouTube does offer separate options for advertisers to transact against Nielsen’s and Comscore’s co-viewing measurements.

“When you talk about co-viewing with other platforms, it’s usually going to be a third-party measurement system that’s going to be providing the data. This is more YouTube grading their own homework, if you will,” said the second agency executive.

A YouTube spokesperson said the company is prioritizing bringing third-party measurement and validation of co-viewing measurement to the market globally ahead of the transaction start date.

There’s also just a general skepticism around co-viewing measurement among some agency executives. Co-viewing measurement isn’t an exact science. Typically, it’s based on a panel of select audience members and then projected across the total audience using what’s called a co-viewing factor — such as 1.6 — to multiply the number of impressions served by the factor representing the average number of people likely watching on the other side of those impressions.

Squishy as that math may be, it’s been acceptable enough for measuring traditional TV ads, which are largely evaluated based on broad age-and-gender categories. But given the level of audience targeting that advertisers are accustomed to on platforms like YouTube, agency executives are unsure how accurately co-viewing measurement will count — or, more to the point, not count — viewers according to a campaign’s targeting parameters. For example, if three people are watching together but only one is part of the advertiser’s defined target audience, then ad buyers don’t want to be charged for the two other viewers.

“The tricky part is the tracking. If two people are sitting on a couch watching and there’s four [people] in the household, how do you know which of those individuals is actually watching?” said a third agency executive.

Also tricky is how co-viewing measurement should affect ad prices. “Co-viewing is one thing, but how you price it is another,” said the second agency executive. For example, since co-viewing would effectively increase the number of impressions counted for a campaign, it could create a scenario in which advertisers are paying more money while, in actuality, serving the same number of ads. “If we start getting into the ‘Well, we’re not going to bring the CPM down as a much as we include co-viewing,’ then you start to have a fight on your hands,” said this executive.

Ultimately, though, agency executives acknowledge that co-viewing measurement will become a standard part of the CTV ad market. And there have been efforts to bring co-viewing to CTV. Hulu, NBCUniversal and Roku, for example, provide co-viewing measurements through Nielsen. And in a study conducted by Nielsen and commissioned by Google, the measurement firm that 26% of YouTube’s CTV viewership occurs with multiple adults in the room watching together, compared to 22% for traditional TV viewership, according to the YouTube spokesperson.

Given that, the agency executives see YouTube’s postponement as an opening to revisit how co-viewing should be counted in a streaming context and not just with respect to YouTube.

“As measurement evolves, we have to go back and relook at how we did co-viewing [historically] and where does that need to evolve?” said a fourth agency executive.