Michael Kaminsky - Modern Marketing Mix Modelling - How to make wise choices - S3 Ep21 (part 2 of 6)
Brain-hurt scale 10/10 - Marketing Mix Modelling or MMM for short is, to paraphrase Zoolander, 'So Hot Right Now'. It seems every day I see another firm pop up with an AI, ML-powered MMM product that claims to be the silver bullet that will fix our iOS14-induced digital attribution void and transport us back to measurement nirvana. But, as you'll find out by listening to this episode, Caveat Emptor. In order to make wise choices in this area, it's first necessary to understand some of the basics of MMM. What it is and how it works. What it does, what it doesn't do. It's strengths and weaknesses. How to set it up. What data-sources and metrics to use. How to optimize and improve them over time. And how to judge one MMM vendor from the next. As a Reforge fan, I stumbled across Recast and again recommended them by Mike Taylor. This wasn't a planned interview, so you'll have to excuse the poor quality, but I trust the depth of the conversation will make up for it. And I found it so interesting that I thought I'd share it with you all. It should give you a really good crash-course in the world of MMM and help you make wiser choices if you're in the market to implement one of these systems in your organization.