This repository presents the theory and implementation of one of the latest Marketing Mix Modeling (MMM) approaches, which utilizes Carryover and Shape Effect modeling. It comprises two distinct sets of files, including:
-
Carryover and Shape Effect Theory (01_theory)
- Examination of the paper by Jin et al. (2017)
- Introduction of the carryover and adstock method for MMM
- Manual calculation of MMM using their method in Python
-
Utilizing Lightweight-MMM for Implementation (02_lightweight-mmm)
- Guide on implementing calculations with the Lightweight MMM library
- Steps from preprocessing to model fitting and optimization
- Utilization of a real dataset sourced from a Github repository