Customer Churn Analysis
A payment processing firm wanted to analyze its customers’ sales activity and demographic data to test its hypothesis on customer churn rate. The firm was interested in identifying possible trends that could predict the likelihood of churn (Customer churn refers to when a customer ceases its relationship with a company).
- Apply data science to model client segmentations based on demographic data
- Apply logistic regression analysis to test hypothesis on possible churn predictors
- Implement churn prediction model based on results of the logistic regression analysis
- Create an on-line dashboard that buckets customers by ‘risk tier’ to track their probability to churn, segmented by various demographic data
Identified a number of sales activity trends in conjunction with customer demographics that acted as strong predictors for likelihood of churn
On-line churn probability dashboard provided an actionable starting point for the firm to implement a multi-faceted customer retention campaign