IMPLEMENTASI METODE SINGLE EXPONENTIAL SMOOTHING UNTUK PREDIKSI PENJUALAN SEMEN PADA CV DUTA BANGUNAN BERBASIS WEB

Abstract

The fluctuating demand for building materials presents a significant challenge in inventory management and business strategy. CV Duta Bangunan Mandiri, a company engaged in the sale of building materials, requires a reliable forecasting method to accurately project future market needs. This study implements the Single Exponential Smoothing (SES) method for sales forecasting, as it assigns greater weight to recent data and is suitable for datasets without clear seasonal patterns. Several smoothing parameter (α) values were tested to achieve the lowest forecasting error. The results show a Mean Absolute Error (MAE) of 513.63 and a Mean Absolute Percentage Error (MAPE) of 4.57%, indicating high accuracy. The sales forecast for January is 10,136.77 units, providing a realistic projection of potential demand. This information enables the company to develop more targeted procurement and distribution strategies. The application of the SES method proves effective in minimizing the risk of overstocking and understocking, while improving operational efficiency. These findings can serve as a reference for similar businesses in adopting forecasting methods to support more effective and data-driven business planning.