Optimization of Sales Turnover Forecasting with the SARI-MAX Model Approach to Predict Revenue Trends Case Study: Sun Thai Tea

This study aims to apply the SARIMAX model to forecast the daily revenue of Sun Thai Tea by incorporating external variables such as weather and promotions, and to compare its performance with the SARIMA model. The daily sales data used exhibit a weekly seasonal pattern and are influenced by external factors, making SARIMAX a suitable choice due to its ability to accommodate exogenous variables. The SARIMAX model employs optimal parameters obtained through grid search, with an order of and a seasonal order of , resulting in a Mean Absolute Percentage Error (MAPE) of which is more accurate compared to the SARIMA model’s MAPE of The findings indicate that SARIMAX, when configured with appropriate parameters, can provide more accurate daily revenue predictions and support more efficient operational decision-making.