What is Churn? Churn is a measurement of the percentage of accounts that cancel or choose not to renew their subscriptions. A high churn rate can negatively impact Monthly Recurring Revenue (MRR) and can also indicate dissatisfaction with a product or service. Understaing the factors that are causing this situation is crutial and can help the company to adopt certain measures that will decrease the churn rate.
Source: Lexer
One of the ways of analysing this is through Data Science/Data Analysis. Understanding the correlations and casualities among customers and services offered that lead to cancellations allied to predictive models can help the company develop solutions to avoid potential churn customers to seek services of company's competitors.
This code presents a churn rate predictive analysis for a fictional company Let's Talk, that offers internet and phone services. The dataset and web application was provided by Let's Data team after a Data Science workshop. The predictions were done using LogisticRegression Classifier since it's a multiclass classification problem.
Main analysis was done in Jupyter Notebook. The model then was deployed in a Web Application that would serve to the company describe a client to check if it's a potential churn case.