Artificial intelligence isn’t fundamentally rewriting media strategies just yet, but it is certainly speeding things up.
AI is boosting existing goals like personalization, accelerating the customer journey through the marketing funnel, and enabling these efforts at a greater scale.
“I think we’re all still in the learning phase,” Liane Nadeau, EVP, head of media investment, Digitas North America, told Beet.TV in this video interview. She suggested AI is currently “an accelerant of the things that we were already doing,” enhancing rather than replacing established approaches to media planning, audience definition, targeting, and measurement.
Moving measurement from ‘or’ to ‘and’
Digitas is guiding clients away from relying solely on rudimentary metrics like clicks and conversions toward a more holistic view of performance. This involves incorporating measures such as incrementality, multi-touch attribution (MTA), and media mix modeling (MMM). Nadeau emphasized these techniques themselves aren’t new, but the approach is evolving.
“We’re really working to evolve the measurement strategies from an ‘or’ to an ‘and’,” Nadeau explained. While near-term indicators like impressions, reach, and CPMs retain value for campaign guidance, the crucial step is linking them definitively to longer-term success metrics.
“The brands that are moving fastest on this journey are the brands that have the ability to move against multiple metrics,” Nadeau said. Success lies in using a “multitude of metrics, be it a composite score or multiple conjoined metrics into one ecosystem that allows us to make decisions more holistically.”
Trusting, but verifying AI tools
Integrating deep learning tools for campaign optimization, particularly mid-flight, Nadeau suggests a cautious approach. She stressed the importance of “trusting, but verifying the sources,” whether they are third-party data providers or tools embedded within DSPs and other buying platforms.
“We’re especially scrutinizing those of the walled gardens, to be able to leverage the power of what we know they can bring in terms of optimization, by way of AI – but ensure we put guardrails around it,” Nadeau stated.
She said she does not want to “hand over the keys” to platforms whose objectives might not fully align with the advertiser’s or could involve “conflating priorities.” This cautious verification aligns with recent University of Bristol research exploring AI’s integration into creative industries, which notes the importance of human oversight to mitigate potential AI inaccuracies.
Nadeau said her aim is to “tap into the power of it in very precise ways” without relinquishing strategic control.
Balancing data risks with creative potential
A common misstep Nadeau identified when clients embrace new approaches is under-resourcing the tests. “One of the things that I’ve seen clients misstep on when they’re starting to test into new territories is not giving it enough budget, and not giving it enough runway to run,” she warned.
Nadeau predicted the most significant shift in performance marketing over the next few years will come from AI’s application to creative itself. The development of agentic multimodal AI frameworks, designed for hyper-personalized advertising, points towards this future.
“What is going to be truly an unlock for brands is when they bring that power of that technology into their creative,” she concluded, envisioning AI enabling a “content engine that continually feeds performance of the brand.”
You’re watching Beet.TV coverage of POSSIBLE 2025. For more videos from this series, please visit this page.
