Customer Potential Index for Production Program Optimization in Manufacturing SMEs

Abstract

This study proposes a methodology to maximize income using a novel Customer Potential Index (CPI) to optimize production programs in small and medium-sized enterprises (SMEs) engaged in customized manufacturing. The research addresses the gap in existing systems that fail to integrate long-term customer value into production planning. Using regression analysis of accounting reports from 33 clients, we developed a CPI model to predict repeat order likelihood. This index was integrated into a mixed-integer linear programming framework that jointly maximizes current marginal profit, projected CPI-based revenue growth, and minimizes opportunity costs from order postponement. In a real-world case with 11 competing orders, the CPI-based optimization identified a high-potential client with a predicted CPI for a higher-margin but low-potential order, resulting in a better strategic allocation of constrained resources. The model was tested using the root mean square error (RMSE) and mean absolute percentage error (MAPE) metrics and was well represented. The novelty of this research lies in bridging customer analytics and production scheduling without proprietary customer relationship management data and using artificial intelligence (AI), while making advanced planning accessible to resource-constrained SMEs.