This project aims to build a customer lifetime value (CLV) model for a telecommunications company using parametric models. The first step is to build accelerated time failure (ATF) models using different distributions: Weibull, Log-Normal, Log-Logistic, Exponential and Generalized Gamma. The models are compared based on various metrics, and the best model is selected as the final model for CLV calculation.
After selecting the final model, CLV is calculated per customer using the model. The CLV is explored within different customer segments, and insights are derived from the analysis.
The project uses Python and various libraries such as lifelines, pandas, numpy, and matplotlib for data analysis, model building, and visualization. The Jupyter Notebook file contains all the code and explanations for each step of the analysis.
Overall, the project provides a comprehensive approach to building a CLV model using parametric models and exploring the insights gained from the analysis.