Machine Learning in Business: Enhancing Inventory Stability via Forecast-Based Inventory Strategies Using Holt's Linear Trend Double Exponential Smoothing

Ineffective inventory management can lead to stock shortages, affecting sales and customer satisfaction negatively. To address these issue, effective and accurate strategies are crucial. This study aims to enhance inventory management at the Terang Teknik Electronics Store in Pontianak, West Kalimantan, where procurement decisions currently rely on owner intuition rather than systematic analysis. To optimize the calculation of safety stock and reorder point, this study integrates safety stock and reorder point strategies with Holt’s Linear Trend Double Exponential Smoothing forecast method. Sales data from 4,228 items sold between January to December, 2023 were analyzed. Using purposive sampling to select 982 items from the highest sales category “Warranty Lamp” and focusing to MYLED Beta brand. The results include the determination of safety
stock, and reorder point for various MYLED Beta wattages. The safety stock for this product are 5 watts at 4 units; 10 watts at 7 units; 15 watts at 2 units; and 20 watts at 5 units. The reorder points are 5 watts at 8 units; 10 watts at 13 units; 15 watts at 7 units; and 20 watts at 8 pcs units. Implementing safety stock and reorder point strategies, inventory management become more efficient and enhance product availability.