The increasing complexity of product distribution and sales activities at PT Mitra Bersama has created challenges in accurately classifying product performance, particularly due to manual data processing that is inefficient and prone to error. To address this issue, this study aims to develop an intelligent decision-support system capable of classifying best-selling and non-best-selling products using a data-driven approach. The CRISP-DM methodology was applied to guide the overall analytical process, consisting of business understanding, data understanding, data preparation, modeling, evaluation, and deployment. The C4.5 algorithm was used to perform the classification through entropy and information gain calculations to determine the most influential attributes. The results show that Type of Food has the highest information gain (0.0113340), followed by Initial Stock, Unit, Month, and Ending Stock, indicating that product characteristics and early inventory levels play a significant role in predicting sales performance. These findings were implemented into a web-based application to facilitate real-time classification and assist decision-makers in optimizing inventory planning, distribution strategies, and sales forecasting. This research contributes to improving organizational efficiency by providing a systematic, accurate, and accessible tool that supports better strategic decision-making in product sales management.