<![CDATA[In the manufacturing sector, accurate demand forecasting is essential for effective material planning and inventory management. PT. XYZ, a company specialising in the production of corrugated carton boxes, currently faces challenges aligning raw material procurement with market demand due to the use of subjective, non-systematic forecasting methods. This research proposes applying statistical forecasting techniques to develop a more reliable and automated forecasting system. The study utilises historical monthly sales data collected over a one-year period, which are analysed using time series forecasting methods. The models are assessed based on key forecasting error metrics, including mean absolute deviation, mean squared error, and mean absolute percentage error. The model construction, data processing, and visualisation, thereby improving efficiency and reducing manual intervention. The findings reveal that combining seasonal statistical models with programming tools enhances forecast accuracy and supports data-driven decision-making within the organisation. This forecasting system can assist the planning division of PT. XYZ is optimising raw material allocation, reducing excess inventory, and preventing material shortages. In conclusion, the study recommends that PT. XYZ implements the decomposition forecasting model as a practical solution for improving the quality of its sales data. The research contributes to the development of forecasting systems tailored for industrial environments with fluctuating, seasonal demand.]]>
