Predicting Customer Lifetime Value in E-Commerce: A Data-Driven Approach to Enhance Customer Retention Strategies

Predicting Customer Lifetime Value (CLV) is one of the most critical tasks that businesses undertake in order to improve customer retention and optimize marketing strategies. The present paper proposes a predictive model for CLV using several machine learning techniques. The features considered include customer demographics, purchasing behavior, and transaction history. All these features were preprocessed using feature engineering and missing value handling. Several regression models, including Linear Regression, Decision Tree, and Random Forest Regressor, were compared in terms of predictive performance. The study includes the decision trees, hash tables as advance data structures such that they can handle large datasets effectively. Of the ones tested, Random Forest Regressor proved to be superior to the others, having the lowest Mean Squared Error (MSE) and highest R² score, thereby validating its robustness in identifying the complex patterns within the data. The work highlights the time and space complexities of different models. This study contributes to the application of machine learning in customer analytics and provides a foundation for future research to integrate more diverse data sources and enhance model performance through hyperparameter optimization. Future work will also explore real-time prediction capabilities and the integration of multimodal data to further refine CLV forecasting.