Forecasting–Inventory Optimization Model: Integrating Exponential Smoothing with Min–Max and Blanket Order Systems For SMEs

Purpose –The research integrates demand forecasting using the Exponential Smoothing (ES) method to develop an adaptive and data-driven framework for cost optimization in volatile demand conditions. Methodology – A quantitative–descriptive and analytical approach was adopted by combining forecasting accuracy analysis with cost comparison modeling. Two forecasting models—Moving Average (MA) and Exponential Smoothing (ES)—were tested using 2021–2023 demand data. The most accurate model (lowest MAPE) was used to simulate inventory performance through the Min–Max and Blanket Order systems. Sensitivity analysis with ±10% demand variation was conducted to evaluate model robustness, while correlation testing validated forecast accuracy against actual demand. Findings – The Exponential Smoothing model achieved superior predictive accuracy (MAPE = 0.883%) compared with the Moving Average model (MAPE = 1.338%). The Min–Max Stock system produced lower total costs—IDR 116,269,920 (2021), IDR 123,260,400 (2022), and IDR 128,466,720 (2023)—compared with the Blanket Order system, which recorded higher and more volatile costs across the same period. The hybrid Min–Max–Forecasting approach demonstrated higher stability under demand fluctuations and improved procurement efficiency, achieving an estimated 30% cost reduction. Practical implications – This study offers SMEs an evidence-based strategy for integrating forecasting accuracy into inventory control, supporting cost reduction and production continuity in resource-constrained environments. The model can be adopted as a reference for developing adaptive inventory policies within the Indonesian SME food sector. Originality– The originality of this study lies in its hybrid integration of Exponential Smoothing forecasting within comparative Min–Max and Blanket Order frameworks, offering empirical validation for forecasting-driven inventory decisions at the SME scale. The approach provides both theoretical advancement and managerial relevance by aligning predictive accuracy with inventory cost optimization in volatile market contexts.