Third-party measurement of ad exposure will disappear at the same time as broadcast TV, and at some time during the broadcast-to-streaming transition the media industry will adopt a model where the initial audience count will be based on media owner ad server log data, which is then backed by independent verification and proof of co-viewing as separate measurement exercises. This is the opinion of Chris Kelly, CEO at Upwave, who believes we could see the widespread adoption of media owner measurement as the foundation for trading within as little as five years.
Upwave is an analytics platform for brand advertising. It provides ways to optimise brand campaigns, measure brand lift and forecast which homes could become potential brand customers, partly by using surveys with exposed and non-exposed households. Kelly believes the television industry spends too much time today focusing on currency discussions (especially the new and alternative currencies) and not enough time thinking about outcomes, but he also has strong views about the future of measurement in general.
He predicts that five years from now, when streaming has grown in strength and ‘legacy’ linear TV is less powerful, the TV industry will agree to give up on third-party exposure measurement and start to accept media owner data, albeit with third-party data auditing and additional work to show demographics and how many people were watching a streaming session.
His philosophy is simple: third-party exposure measurement was an inherent part of the broadcast television world because without it, nobody knew who tuned in. But streaming gives us ad server log data. Third-party measurement is therefore a legacy requirement, but one which must be supported while broadcast linear TV still carries considerable value. As we move closer to post-broadcast, we can safely ditch it.
Expanding on this point, Kelly says: “It is a fairly subversive view, but the industry is always talking about third-party entities affirming that people saw an ad, when media companies already know this from their ad server data. They know how many times an ad has been shown. This [the media owner as the source of trusted measurement] has to be the future.
“You need the co-viewing factor because a [streaming] media owner will not know if two or five people were in front of the TV, and they may not know demography, and these can be provided as discrete services from a measurement company, but they can be decoupled from the counting.
“Every conference in this industry is talking about counting and currency, and how we need new companies to tell us how many people saw an ad, but in digital you don’t need a third-party. We can audit to ensure there is no systematic under-counting or over-counting. We grew up in a world where nobody knew how many people saw a broadcast ad, but that does not apply in a streaming world. The industry has not digested this shift yet.”
Kelly adds: “There are functions where we will need third-party measurement, but there is a difference between asking for the first draft of the count to be external rather than internal. Conducting the first draft of the count internally is not conventional wisdom yet.”
His prediction is that in five years’ time, media owners will publish the first count of how many people saw an ad that was served in their programming. “I believe this is how it will shake out. It requires that advertisers and media sellers agree to the change, so it will take time because today there is still substantial revenue attached to non-streamed viewing.
“[Broadcast] linear TV is still a big deal in 2023, but it is not the future. Media owners have made big bets in digital and have a foot in both camps today, and people are not comfortable yet with giving up third-party counting.”
Kelly says we will always need third-party counting for broadcast linear TV, and that broadcast could last longer than expected, delaying his prediction.