Case Study: Customer Retention - Predictive Analytics
Background: A key US Telecom client is facing increased customer churn rate in the last 2 years. The telecom market in the US is a saturated market. Customer retention is as important as acquiring new customers.
Approach: All the key factors that could potential influence customer churn were identified. A series of univariate, bivariate analysis was performed to create churn segments. Then statistical modeling techniques like Logistic Regression was used. Decision Tree approach using CHAID helped idenitify pockets of high & low customer churn
Impact: Client now has a strong feel for that factors cause customer churn. Client is now taking actions on the recommendations on potential ways to arrest churn