MENU

Fun & Interesting

A Bayesian Approach to Media Mix Modeling (Michael Johns & Zhenyu Wang)

PyMC Developers 33,245 4 years ago
Video Not Working? Fix It Now

Speakers: Michael Johns & Zhenyu Wang Title: A Bayesian Approach to Media Mix Modeling Video: https://youtu.be/UznM_-_760Y Event description: This talk describes how we built a Bayesian Media Mix Model of new customer acquisition using PyMC. We will explain the statistical structure of the model in detail, with special attention to nonlinear functional transformations, discuss some of the technical challenges we tackled when building it in a Bayesian framework, and touch on how we use it in production to guide our marketing strategy. Discourse Discussion https://discourse.pymc.io/t/a-bayesian-approach-to-media-mix-modeling-by-michael-johns-zhenyu-wang/6024 ## Outline - Marketing at HelloFresh - What is a Media Mixed Model? - Media Mixed Model: Structure & Function - Bayesian model building workflow - Model applications ## Timestamps 00:00 Introduction and outline of presentation 01:16 Marketing at HelloFresh (funnels, conversion, channels) 02:40 Measuring the effectiveness of marketing 05:00 What is Media Mix Modelling? (Multivariate regression model) 06:20 Structure of a Media Mix Model 07:51 Transformation functions (Reach function and Adstock function) 10:53 Benefits of using Bayesian methods to build a Media Mix Model 13:07 HelloFresh's Media Mix Model structure 19:46 Geometric Adstock Function 20:54 Nonlinear Saturation Function 21:16 The Bayesian MMM workflow 22:39 Applications of HelloFresh's Media Mix Model 26:41 Constrained optimization algorithm 29:18 Thank you! Speaker info: Michael Johns is a data scientist at HelloFresh US. His work focuses on building statistical models for business applications, such as optimizing marketing strategy, customer acquisition forecasting and customer retention. LinkedIn: N/A Zhenyu Wang is a Senior Business Intelligence Analyst at HelloFresh International. He works on developing and implementing methods to measure the effectiveness of advertising campaigns using analytic and statistical methods. LinkedIn: N/A Part of PyMCon2020. More details at http://www.pymcon.com #bayesian #statistics #marketing

Comment