In the competitive banking sector, accurately predicting customer subscriptions to term deposits is crucial for optimizing marketing strategies and maximizing revenue. However, traditional models often struggle with both predictive performance and explainability. To overcome these challenges, we propose an advanced approach that enhances accuracy while ensuring model transparency. Our method integrates the CatBoost ensemble model renowned for its efficiency in handling categorical data with SHAP (SHapley Additive exPlanations), a leading Explainable AI (XAI) framework. By leveraging SHAP, we provide clear, interpretable insights into model predictions, identifying key factors influencing customer decisions. SHAP quantifies each feature’s contribution, ensuring reliable explanations that enhance trust and accountability in financial decision-making. This approach not only improves predictive accuracy but also harnesses the efficiency of the tree-based SHAP explainer to enhance transparency while significantly reducing computational time. By offering an interpretable and efficient predictive framework, our method supports informed decision-making with minimal computational overhead, making it highly valuable for bank term deposit predictions.