Can't sleep? Great, join us for the data science midnight session!
=== Streaming is starting @ 12:30am ===
=== Small intro ===
=== Talk 1 ===
Luca Fiaschi VP Data hellofresh on "Bayesian Media Mix Modeling using PyMC3, for Fun and Profit"
Media mix models (MMM) are multivariable regression models used to quantify the effectiveness of advertising on a key business metric as new customers acquired. Typically, marketing dollars spent each week on specific campaigns (think Hulu, Facebook, mailers, banner ads, etc.) are used to predict the total number of new customers acquired that week. MMMs are becoming an attractive option for businesses given the limitations of contemporary digital measurement methods. Jin et al. (2017) recently proposed using a Bayesian approach to build an MMM with transformations on marketing activity variables (e.g., spending) to account for diminishing returns and decaying effects of ad exposure over time. This talk will describe how we built a Bayesian Media Mix Model of customer acquisition using PyMC3. 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. This talk should be interesting to general community members with an applied focus as well as members in the industry using PyMC to solve common business problems.
=== Break & Networking ===
=== Talk 2 ===
Michael Schneider, Data Scientist @ Shopify on "Uplift model and experimental setup - Das Sauerkraut of measuring the true effect of your target. "
=== Closing ===