Marketing Mix Modeling – There’s Danger in “Knowing Enough to be Dangerous”
Marketing Mix Modeling (MMX) has become deeply rooted in marketers’ decision-making processes, but understanding of the tools’ capabilities remain superficial at best. In this webinar, Bruce Pivarunas will review the historical strengths of MMX as well as the statistical limitations that — if not addressed — can cause marketers to unknowingly make unsound ROI-based decisions.
This webinar will visually bring-to-life the causes of MMX’s statistical dangers -- provide real-world examples of how marketers have inadvertently fallen into these hidden traps – and propose ways for MMX modelers and end-users to “reconnect” with the core mission of the service by thinking beyond statistical parameters and functional forms to the blend of consultative art and science that’s needed to create great marketing.
Attendees will learn:
- Why MMX vendors being “personally objective” is not the same as their being “statistically unbiased”
- How to clear the distortions that come from viewing “today’s personalized continuity marketing” through “yesterday’s mass-market near-term focused lens”
- Why “statistically controlling” for a variable (seasonality, trend, etc.) does NOT mean removing its influence on marketing performance