<![CDATA[XYZ Company, a local perfume company in Indonesia, repeatedly faces problems in their monthly operations, such as stockout and overstock issues due to fluctuating demand and uncertain production lead times from manufacturers. This phenomenon risks reducing product availability, hindering cash flow, and lowering consumer trust and experience. To address this situation, the study evaluated four time-series forecasting methods (3-Period Moving Average, 5-Period Moving Average, Simple Exponential Smoothing, and Holt’s Double Exponential Smoothing) using sales data from the year 2024 until early 2025 and measured their accuracy using MAD, MSE, and MAPE. The results showed that the Simple Exponential Smoothing method produced the lowest error (MAD 185,92; MSE 78.207,25; MAPE 12,53%), making it most suitable for the characteristics of fluctuating short-term demand. Based on the forecast, XYZ company is advised to set a safety stock of 613 units and a reorder point of 2.109 units to maintain availability during lead time. Further recommendations include forecasting per variant category, consideration of marketing activities in demand projections, and negotiation of lead time commitments with manufacturers as risk mitigation.]]>
Copyrights © 2025
