Conducting research on MMM: Introduction
Adopting MMM is far too difficult today - we want to make it easy!
This article is part of Sellforte's MMM Procurement guide. We are in chapter 1B of the guide, as illustrated below:
At the start of the MMM journey, three things are typically considered:
Understanding the journey ahead: It is important to understand the steps that are typically needed to implement MMM. For bigger companies it might take several months to get there, but for small and agile companies, you might be able to fast-track the process and get to results very fast
Conducting research on MMM (focus of this article): It's important for you to know the basics of MMM, so that you can credibly talk about it with your stakeholders
Identifying your stakeholders: It's not a one-man journey. It helps if you can early-on identify the people in your organization who you need to bring with you for the journey
In this article we will share sources where to find information on MMM.
1. Sellforte Marketing Mix Modeling guide
Sellforte Marketing Mix Modeling guide provides a good overview to MMM, It has has both simple high-level overviews, such as intuitive example how MMM works, as well as technical deep-dive articles, such as what is adstock.
2. Research articles by Google and others
For data scientists looking to dive deeper into MMM methodology, we recommend following research papers. Google Research has been the most active in the field:
Challenges And Opportunities In Media Mix Modeling (2017, Google - Chan et al.)
Bayesian Methods for Media Mix Modeling with Carryover and Shape Effects (2017, Google - Jin et al.)
Geo-level Bayesian Hierarchical Media Mix Modeling (2017, Google - Sun et al.)
Hierarchical Bayesian Approach to Improve Media Mix Models Using Category Data (2017, Google, Wang et. al)
Bayesian Time Varying Coefficient Model with Applications to Marketing Mix Modeling (2021, Uber - Ng et al.)
Hierarchical Marketing Mix Models with Sign Constraints (2020, Cheng et al.)
Bayesian Hierarchical Media Mix Model Incorporating Reach and Frequency Data (2023, Zhang et al.)
Media Mix Model Calibration With Bayesian Priors (2024, Zhang et al.)
3. Google Meridian and Meta Robyn docuementation
Google's and Meta's public MMM libraries include also a short introduction to MMM, as wells as deep-dives to select technical topics they consider important: