Selection of the Best Forecasting Method for Newspaper Demand Based on Special Event (Case Study: PT. XYZ).

The development of technology and information has a significant impact on society, both positively and negatively. Digitalization has resulted in many people switching from printed newspapers to digital newspapers. This could lead to demand uncertainty in the newspaper industry. At PT.XYZ, the number of newspapers produced and distributed to realtors often exceeds customer demand. In 2017, the company's return rate reached 5%. However, every January, March, and April where the Chinese New Year Festival, Dalang Cilik Festival, and World Dance Day are held, the demand will increase. This surge in demand, if not anticipated properly, can result in unfulfilled customer demands. For this reason, in this research, forecasting is carried out by considering special event factor using Holt-Winter's, Decomposition, and Exponential Smoothing Event-Based method. To decide which method is the best, the MAPE value was used as criteria. The MAPE value for Holt-Winter's, Decomposition, and Exponential Smoothing Event-Based method respectively are 0,92%, 1,05%, and 3%. Based on the MAPE value for each method, it can be seen that the most effective forecasting method is Holt's-Winter method with a MAPE value of 0.92%. [ABSTRACT FROM AUTHOR]