YouTube, which will begin trading based on its own surveys to calculate how many people are watching ads on connected TV, is facing opposition from some agencies who see it as a move away from third-party measurement by the platform.

Also read: How YouTube is plotting the future of video

The change, announced to clients earlier this month, will be implemented in January and will impact YouTube Select video inventory on CTV. YouTube has, for some time, used surveys that ask viewers how many people are watching on the big screen to help develop a co-viewing multiplier. Since 2022, the unit of Alphabet’s Google has reported that data to advertisers for planning purposes only. Starting in January, it will begin billing based on its co-viewing data, said Kate Alessi, managing director of YouTube/Video Global Solution. The move is global, though it will be implemented differently in markets depending on the measurement partners YouTube uses.

Advertisers will still have the option of negotiating guarantees based on Nielsen Digital Ad Ratings and its panel-based co-viewing data, or via Comscore Campaign Ratings and its co-viewing estimates, Alessi said. 

Nielsen has no plans to start using co-viewing survey data from Google to inform audience measurement in the U.S., according to a spokesman. But the company does, in some cases, use surveys it administers itself to help measure co-viewing in some markets. 

Why co-viewing matters

Watching YouTube on mobile or desktop devices is usually a solo endeavor, but watching it on the big screen in living rooms is another matter. A study earlier this year by iSpot.tv-backed measurement firm TVision found YouTube CTV has an average of 1.48 viewers per household, slightly better than some others such as NBCUniversal’s Peacock (1.41) but well behind Disney+ (1.71). The number varies widely by day part, programming and family structure, as primetime and households with children and children’s programming tend to have more people watching. 

Also: Paramount and Omnicom test Nielsen buy-side alternatives

Co-viewing has always been part of TV audience measurement. Nielsen's ability to estimate co-viewing through its panel of 41,000-plus households has been a key part of its advantage over measurement rivals that rely on data from set-top boxes or smart TVs, but use smaller panels to estimate how many people per household are watching.

Adding a strong co-viewing number can boost audiences and revenue significantly (about 48% for YouTube CTV over its mobile or desktop audiences, based on TVision’s estimates). That’s why advertising executives, speaking on background, believe Google is pushing to maximize its co-viewing multipliers now in the face of slowing ad revenue growth in the past year, and why they’re suspicious of the move.

Agency pushback

“The fundamental problem is that it contradicts their approach and the principles they laid out I think a couple of months ago, when they called for third-party measurement and called for it to be comprehensive,” said one agency executive. “But then they’re doing their own thing and pushing that down in the marketplace.”

The push is soft at this point, given that advertisers can opt for Nielsen or Comscore numbers instead. That doesn’t rule out YouTube making its first-party survey data mandatory someday. 

“As of today, we believe we should give that choice to our customers, and so that is our intent, to continue to offer the Nielsen DAR guarantees,” Alessi said.

So why would advertisers opt for YouTube’s own data rather than that of Nielsen or another third party?

“Ultimately customers make choices for a variety of reasons,” Alessi said. “We offer Nielsen DAR guarantees on YouTube Select today, and we see some customers choosing to transact on a first-party basis, and others choosing to transact on a third-party basis. Every customer has different reasons.”

Some agency executives saw the communication they received from YouTube’s sales team as an effort to force acceptance of the platform’s first-party co-viewing data, but Alessi said the agencies would have a choice. It’s not yet clear what the pricing implications for clients will be in opting for Nielsen guarantees vs. trading on YouTube’s first-party co-viewing data, agency executives said.

Widespread acceptance of YouTube’s data would set a bad precedent for the marketplace, said one agency executive, because it would open the door for every other streaming platform to set up its own survey-based system to get a co-viewing number.

“If I’m at NBCU or Disney or whatever, I now have the ability to create my own factor and they would be able say if we let Google do it, why not let another group do it?” the executive said. “It becomes a slippery slope. It sets an entire industry precedent, and we are pushing against it, because there’s always a benefit for third-party validation of anything.”

Methodology questions

But at least one measurement company executive who has looked at YouTube’s methodology for counting co-viewing believes it looks valid.

The platform describes it in general terms as a complex process. A YouTube primer on co-viewing describes it this way: “Google measures co-viewing on TV screens using a combination of Google's connect TV panel data and real-time, census-level surveys to assess co-viewing behavior across 100+ countries via 70+ languages worldwide. Statistical methods are used to eliminate bias and to make sure that the results accurately represent YouTube’s viewers.”

In other words, it's more than just pop-up surveys asking people watching YouTube on TV screens how many people are watching, Alessi said.

While Google describes the process as including “census-level surveys,” not every viewer of YouTube content gets served with a survey on every viewing occasion. A video YouTube made public in 2022 about how it measures its own co-viewing went behind a restricted wall earlier this year.

Nielsen also has declined to provide much public detail about how it calculates co-viewing for YouTube and how that might vary from co-viewing for linear TV.  “As consumer viewing habits have shifted, Nielsen has been diligently working to evolve its audience measurement methodologies to provide clients the most accurate and reliable data possible,” read a statement from Nielsen. “For Nielsen, this means leveraging the significant value of large, digital data sets alongside trusted panel data, which is underpinned by behavioral signals from real people.”

Comscore, iSpot and VideoAmp representatives did not comment on queries about how they will handle YouTube co-viewing measurement going forward.

Amazon also opens new currencies

Up to now, the considerable discussion around alternatives to measurement heavyweight Nielsen has come mainly from TV networks such as NBCUniveral, Paramount, Warner Bros. Discovery, Fox and TelevisaUnivision. 

But YouTube’s move to offer its own alternative currency follows Amazon Prime’s recent move to open the door for trading on alternatives such as VideoAmp and iSpot.

Read more about Amazon Prime here

Neither platform is attributing those moves to dissatisfaction with Nielsen, but agency executives believe that’s what it’s about. Both YouTube and Amazon are spending big money on sports programming and now need the ad revenue to support it, said an agency executive. “They spend so much money on it that it has to go right,” the executive said.

For YouTube, the only official new currency alternative is its own co-viewing data. But in June, YouTube also moved to begin sharing data through its Ads Data Hub clean room with iSpot and VideoAmp for planning purposes. Should YouTube follow the same progression as with its own co-viewing alternative, making those Nielsen rivals available as trading options would come next.

Amazon announced in May that it will begin working with VideoAmp and iSpot to provide measurement. People familiar with the matter said Amazon is offering the Nielsen rivals as shadow currencies to compare against Nielsen for the 2023-24 season of “Thursday Night Football,” then allow trading on them 2024-25.

That move comes a year after Amazon reported its own server-based data on TNF audiences alongside Nielsen estimates each week, with Amazon’s numbers consistently higher. But an Amazon spokesman said the embrace of VideoAmp and iSpot was not a result of displeasure with Nielsen, and like YouTube couched the move as offering advertisers more choice.

“We always work back from customer needs and wants to determine the services we offer, and we hear from advertisers that they want choice in measurement,” according to an Amazon statement. “Our aim is to provide advertisers with a robust set of measurement and insights from multiple services in order to help them make more informed decisions. We will continue to expand to other service providers as we get additional requests from advertisers.”

CTV measurement battle grows

All of the moves come as advertisers have shown growing interest and concern with CTV measurement as the medium continues to win viewers and budgets from a fast-shrinking linear TV market, according to one measurement industry executive. 

CTV measurement is fast shaping up as a more competitive marketplace for Nielsen than linear TV, and the executive said one future scenario could be where Nielsen maintains dominance in a shrinking linear TV market but multiple currencies become commonplace in streaming.

Much of the attention on this front has gone to the U.S. Joint Industry Committee, which is made up of major TV network companies, and how it will certify measurement companies to get server streaming data aggregated from such participants as NBCU, Paramount, Fox and Warner Bros. Discovery (but not, at this point, YouTube, Amazon or Disney Advertising, among others). But even outside the JIC, big platforms are opening up to alternative measurement, or their own measurement.

Nielsen isn’t participating in the JIC certification yet, and might not be able to get server data from at least some JIC members if it doesn’t. Rivals such as iSpot, VideoAmp and Comscore also appear to be ahead of Nielsen in one-on-one server integrations with streaming networks for their viewing data. And YouTube’s move to offer its own co-viewing data as currency hits at one of Nielsen’s biggest competitive advantages in the U.S.—having the biggest household panel for estimating how many people in each household are watching programs.