Predictive Customer Churn Analysis and Retention Strategy Development Using the Global Superstore Dataset

Customer churn Customer churn is a business concern as it has a direct effect on revenue and customer lifetime value (CLV). In this case, the Global Superstore dataset is used for churn prediction and implementing effective retention strategies through machine learning models. Logistic Regression, Random Forest with Boruta, and XGBoost with a Random Forest Ensemble with SMOTE augmentation are employed to identify at-risk customers for churn. Models are compared based on accuracy, precision, recall, and F1-score. K-Means clustering categorizes customers based on the risk of churn, allowing for personalized retention strategies such as targeted marketing and loyalty programs. A/B testing is used for identifying the effectiveness of these measures and making conclusions for customer retention and profitability enhancement.