Ryan Dew from the Wharton School (University of Pennsylvania) recently led an engaging session at the WU Vienna University of Economics and Business as part of the Marketing Research Seminar Series. He discussed his paper, Your MMM is Broken: Identification of Nonlinear and Time-varying Effects in Marketing Mix Models. His work dives into the limitations of traditional Marketing Mix Models (MMMs) and highlights how confusion between nonlinear and time-varying effects can misguide marketers in their budget decisions.

Dew introduced a fresh perspective by leveraging a Bayesian nonparametric framework with Gaussian Processes, enabling a clearer distinction between these complex effects. "Marketers often struggle with interpreting data correctly, which can lead to suboptimal budget allocations," Dew pointed out. He emphasized the need for experimental methods to enhance understanding of different model types.

In addition to his research presentation, Dew also held a talk for doctoral students, focusing on the challenges of probabilistic machine learning considering the decreasing availability of individual-level customer data. His insights on this pressing issue sparked lively discussions among attendees, shedding light on the future of data-driven marketing.

We thank Ryan for his visit and valuable contributions to the ongoing discourse in marketing analytics!