Clean rooms aren’t going away
The partnership also underscores the continued role of data clean rooms, which some industry observers have suggested could become less relevant as AI adoption grows. Optable and Goodway instead describe AI as a foundational layer—enabling secure data sharing and making sensitive datasets, such as transaction-level retail data, usable within AI systems.
“You can’t have a sustainable LLM without a proper clean room,” said Tom Swierczewski, vp of media investment at Goodway Group. “There’s currently 38 different privacy regulations in the United States alone. They’re only going to be on a rise consistently.”
New skillset
For Goodway, deploying AI agents is also changing how teams interact with data. Rather than navigating dashboards or manually querying datasets, planners are now required to increasingly work in conversational interfaces.
This entails asking “good” questions, refining prompts and iterating on outputs, according to Wolk.
“A conversational environment is a very different experience with technology and data—it’s a new muscle and skill set for practitioners,” Wolk said.
Executives said the technology is expected to improve efficiency over time, though they acknowledged an initial learning curve as teams adjust to new workflows. For now, success is being measured through adoption, time savings and driving sales.
“AI is directional at best—sometimes poorly directional, sometimes reasonably accurate,” Optable’s Dumas said.
That shift is also changing how audience building happens in practice. Where media teams once manually queried datasets to create segments, AI agents are beginning to take on that role.
“In the old world, someone would type in, ‘We want to create a segment of people who bought a bagel in the last 30 days and retarget them,’” Dumas said. “In the agentic world, you’re talking to an agent that sits across all these datasets, organizes them and proposes audiences for you.”
