Abstract

Customer retention has emerged as a strategic priority in retail banking, where high churn rates directly impact profitability and growth. As customer acquisition costs continue to rise, banks increasingly rely on predictive analytics and machine learning to anticipate attrition and design proactive interventions. The paper explores the application of predictive modelling techniques ranging from logistic regression to advanced deep learning in identifying at-risk customers and informing targeted retention campaigns. Predictive models can assign churn risk scores with increasing accuracy by leveraging diverse data sources, including transactional records, demographics, service usage, and digital adoption patterns. The study also examines how prescriptive analytics translates these predictions into actionable marketing strategies, such as personalized offers or loyalty incentives, that align interventions with customer value. The findings underscore the potential of data-driven approaches to enhance competitive advantage, reduce revenue leakage, and foster long-term customer loyalty in retail banking.