Measurement, from exposure to outcomes, is a priority for marketers, especially given the growing complexity of media. Two projects aim to outline solutions and provide support for the development of improved metrics.
Read more »
Member Only Access
In this AI session, Laura Denton Carter, Travis Gossen, and Doug Horn unveiled a set of AI-powered tools developed by Google that streamline the creative development process and enhance targeting capabilities. Paul Donato of the ARF moderated the subsequent question-and-answer portion. The presentation focused on how these tools optimize campaign efficiency and creative performance. Real-world use cases illustrated how AI can drive personalization, audience segmentation, and dynamic content creation across platforms. The session also emphasized the importance of ethical considerations and human oversight when implementing AI in advertising.
Member Only Access
Lauren Littlejohn – Director, Data Science & Research, Kroger Precision Marketing
Lauren Littlejohn of Kroger Precision Marketing presented a novel ad measurement solution developed by her team. She discussed the evolution of Kroger Precision Marketing's measurement, audience and optimization capabilities, emphasizing the importance of privacy and the challenges of cross-channel measurement in today's advertising landscape. Lauren introduced Precision View 360, a new product designed to provide granular, cross-channel measurement, while considering privacy concerns and the complexity of multiple media channels. The product leverages machine learning, randomized controlled trials and household matching methods to deliver insights that transform advertising effectiveness.
Key takeaways:
- Kroger captures over 96% of transactions on their loyalty card, providing rich customer data for measurement.
- Clean rooms are used for measurement in platforms like Meta, Disney, Pinterest and Snapchat.
- Household level measurement isolates the advertising effect from other market variables.
- The hybridized model combines household level experiments with a Bayesian marketing mix model.
- The model provides a comprehensive report on the total incremental effect of media plans.
- Interaction effects between different channels are quantified to understand synergistic impacts.
Watch the Presentation
Download Presentation
Member Only Access
Sergey Fogelson – VP, Head of Data Science, TelevisaUnivision
Pouya Tehrani – Director, of Data Science, Graph & Insights, TelevisaUnivision
Sergey Fogelson and Pouya Tehrani from TelevisaUnivision described their development and utilization of the TelevisaUnivision household graph. This is an innovative tool used to identify and market to Hispanic Spanish-speaking households in the U.S. The graph is built using a combination of first-party and licensed third-party data sources, linking various identifiers at the household and individual levels. Machine learning is then employed to identify the language from each domain held within their large data pool. The graph’s objective is to enhance targeting strategies, improve personalization and identify growth opportunities within the market, by understanding language preferences.
Key takeaways:
- The TelevisaUnivision household graph helps identify Hispanic households likely to consume Spanish language media.
- The graph is built using first-party and licensed third-party data sources.
- It links identifiers at the household and individual levels to target and measure ads.
- The project aims to add language preferences at the household level for better targeting.
- A new dataset of third-party web traffic logs provides finer-grained understanding of media consumption.
- The intelligence from the graph helps personalize content and improve streaming hours.
- The project identifies regional content preferences among Spanish-speaking households.
Watch the Presentation
Download Presentation
Member Only Access