Is the potential for machine learning and artificial intelligence to revolutionize premium video advertising is being held back by an unglamorous problem?
Some in the industry think of a common identity framework is needed to prevent inconsistent reach and skewed measurement.
David Levy, CEO of OpenAP, agrees. In this video interview with Beet.TV, he said the industry must first tackle this foundational “boring work around consistency” before it can reap the rewards of advanced optimization.
Clean rooms offer a path through fragmentation
The persistent challenge of fragmentation has historically forced industry players to build their own bespoke infrastructure, but a significant shift is now underway, according to Levy.
Data clean rooms have emerged as a critical technology, enabling secure data collaboration that was previously difficult due to privacy rules. No wonder the global market for data clean room platforms projected to expand substantially by 2032 as the industry seeks privacy-compliant targeting solutions.
“With the advent of clean rooms and most big companies now having their data exist inside clean room environments, we’re now able to build federated environments that go across those clean rooms,” Levy explained. This allows an advertiser to “take one audience and immediately match that across multiple programmers and actually do planning, measurement, optimization, etcetera.”
The ultimate ambition is to simplify the entire process, breaking down silos to benefit advertisers of all sizes, he added. “How can we get to a place where it’s easy for even a small advertiser to take their audience and find that audience across 50 different publishers all in a long form setting?” Levy asked. “That is now doable and possible with that clean room infrastructure.”
The need for a single source of identity
But, while clean rooms provide the secure environment for collaboration, they do not solve the underlying identity problem on their own. OpenAP recently launched its OpenID solution to directly address this challenge by creating a common identity spine for cross-platform campaigns.
“If you’re an agency and you have one central notion of who you’re trying to reach, and you then share that with multiple different programmers… if they all use disparate notions of which identifiers belong to which households, which they do, ultimately you’re actually targeting different audiences,” Levy said. A post-campaign measurement company with yet another identity graph only compounds the issue, making a campaign appear “significantly out of target.”
The solution, he argued, is to establish an upfront “source of truth” for both the audience and the identity layer used to activate and measure it. “What we need to do is not just share a list of audience IDs in the front of the campaign, but actually have the agency define the list of audience IDs and also the identity that needs to be used on the campaign and what’s going to be considered on target post campaign at the beginning,” he said.
