How Omnicom Media Group is making sense of clean room complexities

One example of clean rooms’ complexity is that, while typically associated with first-party data, a marketer is not required to have first-party of its own to use a clean room, according to Rossen. “For an advertiser or brand, in particular, whether you have first-party data at scale or not is really not the question,” he said. “What the question is is how are you bringing yourself today in a capability, in a position to use clean rooms in your ecosystem for media impact management, optimization and really planning and investment insights as well.”

Those use cases will be contingent on publishers and platforms plugging their own first-party data into clean rooms for advertisers to be able to measure and assess their campaigns’ performance. And while some on the sell-side are still sorting out their clean room strategies, others such as Disney, Google, Meta and NBCUniversal have either rolled out or begun testing clean rooms with advertisers and agencies like OMG.

“Every major supply-side publisher is going to be building out their own clean room environment,” Rossen said. Some will rely on clean room providers, such as Habu, InfoSum and Snowflake, which can facilitate some interoperability among publishers’ clean rooms, while others will opt to build their own proprietary clean rooms.

Of course, the widespread rollout of clean room options complicates matters in its own way. Publishers have already complained about the prospect of needing to support all kinds of different clean rooms, and Rossen acknowledged this complexity.

“The challenge is going to become how do we take all these disparate clean rooms — if you have media dollars in a walled garden ecosystem, media dollars in streaming and TV, media dollars in multiple different places — how are you going to bring that all together to look at a consolidated holistic view of your media investment. [This connection among clean rooms] is the next challenge we have to tackle together,” Rossen said.

But doesn’t the idea of connecting data across clean rooms defeat the entire purpose of having these information-safe houses? Nope, said Rossen. “It doesn’t defeat the purpose because it’s about protection of data rights.”

Rossen explained the clean rooms’ function for protecting data in a way that recalled an analogy of a locked metal briefcase handcuffed to a courier’s wrist in order to secure its custody. In that analogy, the clean room would be the briefcase, its contents would be the data and the courier would be the clean room provider. 

All analogies are oversimplifications, but this one would seem to hold up once there are standards in place that would enable data stored in one clean room to be shared into another clean room while complying with privacy regulations as well as companies’ own privacy policies. When it comes to clean rooms that level of standardized interoperability appears to be the biggest challenge and complexity of all.

“How do you share data in a more privacy-compliant way without any identity at all through using analytics and algorithmic matching, which is a step above where we are today?” Rossen said. He noted that the IAB Tech Lab’s Privacy Enhancing Technologies working group — of which he’s a member — is working on such standardization. “No clarity on what that’s exactly going to be yet, but that’s the direction we’re heading together.”