COMPARISON OF ARIMA AND WINTERS METHODS ON SALES FORCASTING OF FURNITURE COMPANIES AT UD PODOMORO ASAHAN

Abstract

Economic growth in Indonesia affects the need for household furniture. With the increasing demand, consumers have to wait a long time for the requested furniture. This causes sales to not be optimal due to consumer disappointment which results in cancellation of requests. The purpose of this research is to determine the optimal development of sales and make a production schedule based on the forecasting results obtained. The data of this research are window sales at UD. Podomoro Asahan from July 2016 to July 2021. The data was processed using the ARIMA and Winters methods and then compared between the two methods.The results of the study used the ARIMA method, namely using the ARIMA model (0,1,1) with an error value of MPE -0.079772% and MAPE 16.592778%. The results of the study used the Winters method, namely using the smoothing constant α=0.546225; β=0.259846; γ=0.116178; with MPE error value -0.39785% and MAPE 39.78471%.

Keywords

• forcasting
• arima
• winters
• master production schedule.