AI-optimized, omnichannel campaigns are changing how media are planned, activated and measured. Drawing on in-depth interviews with advertisers and providers, the ARF Cross-Platform Measurement Council’s Attribution Working Group examined omnichannel campaign tools such as Google Performance Max and Meta Advantage+, exploring their potential to simplify execution, optimize across channels and improve performance. Their research also identified important tradeoffs, including transparency, testing capabilities and ongoing measurement and governance challenges.
Member Only Access
AI can accelerate campaign measurement, but how do you know the results are reliable? On July 15, the ARF Cross-Platform Measurement Council led an exploration of validating AI-powered measurement models and systems. Industry practitioners shared validation approaches, frameworks, and tools that work based on real case studies. Plus, attendees had the opportunity to ask questions in a Q&A.
Member Only Access
As privacy changes make it harder for marketers to track individual customer journeys, media and marketing mix modeling (m/MMM) is experiencing renewed interest. Open-source tools from Meta, Google and the PyMC ecosystem promise to make these methods more accessible, particularly for organizations without the resources for customized proprietary models. In this MSI working paper, Julian Runge and Koen Pauwels evaluate the leading open-source approaches, examining their capabilities, differences and tradeoffs. Their analysis shows that while automation can democratize sophisticated marketing measurement, using these tools responsibly requires careful attention to model assumptions, diagnostics, calibration and validation.
Member Only Access
Artificial intelligence is becoming deeply embedded in advertising and marketing, but adoption alone is no longer the story. In this guest article for The Advertising Club of New York, ARF Senior Director of Research & Insights Tracy Adams examines findings from the ARF's latest AI research, which reveals that while organizations are increasingly confident in AI-generated outputs, many are still developing the testing, governance and measurement practices needed to ensure those outputs deliver meaningful business value.