Customer Lifetime Value Prediction Using Linear Regression to Support CRM Strategy in Banking Industry

Abstract

This study aims to predict Customer Lifetime Value (CLTV) using a Linear Regression approach to support Customer Relationship Management (CRM) strategies in the banking industry. The research utilizes customer transaction and demographic data to identify key factors influencing customer value. The workflow includes data preparation, exploratory data analysis, preprocessing, model development, and evaluation. Variables such as age, transaction frequency, balance, and product usage are analyzed as predictors of CLTV. The Linear Regression model is selected due to its interpretability and ability to explain relationships between variables. Model performance is evaluated using metrics such as Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE). The results show that several variables significantly affect customer value and can be used to estimate future contributions of customers. The predicted CLTV is then utilized to support CRM strategies, including customer segmentation, personalized product offerings, and retention programs. This study demonstrates that a data-driven approach can enhance decision making and improve service effectiveness in banking. The findings provide practical insights for financial institutions in optimizing customer management and increasing long term profitability. Furthermore, the model supports strategic planning by enabling early identification of high value customers and improving allocation of marketing resources effectively and efficiency