Manual inventory management at UD Ema Kencana Abadi has the potential to cause recording errors and stock imbalances. The primary focus of this study is to implement a web-platform-based inventory system that integrates forecasting techniques as a decision-support instrument. The method used is Single Exponential Smoothing (SES) with dynamic optimization of the alpha (α) parameter through iteration to obtain the lowest Mean Absolute Percentage Error (MAPE), which is then compared with the Single Moving Average (SMA-3). Testing was conducted using nine months of historical sales data on six building material products. The results showed that SMA-3 is more optimal for data with stable fluctuations, yielding a MAPE of 11.91%, whereas the optimized SES is more accurate for data with medium to high fluctuations. The implementation of this system improves forecasting accuracy and supports more objective stock procurement decision-making.
Copyrights © 2026