A demo of my usage of Marketing Mix Modeling in PYMC
Please forgive the mess as I am still working on this project on my own personal time. The mmm.ipynb file contains a notebook where I am working on this project. The assumed business case is that I am working for a Marketing Agency and our client wishes to optimize their ad budget. The first task towards this goal is to analyze which advertizing channels are driving their revenue gains and where they can find the most profitable places for greater investment.
As this is not a finished product, I hope any prospective employers who view this project will take it not as a demonstration of skill in subject so much as proof of an ongoing drive to learn a subject and improve.
[1] Traditional and digital media impact on Sales. (2019). AnandKalidasan. Kaggle. https://www.kaggle.com/datasets/mediaearth/traditional-and-digital-media-impact-on-sales/
@techreport{46001,title = {Bayesian Methods for Media Mix Modeling with Carryover and Shape Effects},author = {Yuxue Jin and Yueqing Wang and Yunting Sun and David Chan and Jim Koehler},year = {2017},institution = {Google Inc.}}
Borden, Neil H. (June 1964). "The Concept of Marketing Mix" (PDF). Journal of Advertising Research: 2–7. OCLC 222909833.