Comparison of Double Exponential Smoothing Method with Weighted Moving Average in Forecasting UD Sales. Setya Abadi D. M as Financial Literacy

Abstract

Purpose: UD. Setya Abadi D. M is a business unit focusing on the food industry, such as making frozen food (Maryam, churros, tortillas). In carrying out its production activities, this company certainly needs a strategy to compete with other competitors, one of which is forecasting. In this research, the data used is past product sales data from July 2021 to June 2023. This research aims to forecast sales in the future period in order to meet customer needs.Methodology:This data processing uses the weighted moving average and double exponential smoothing methods, which are then tested with the Mean Absolute Percent Error value. Forecasting with the lowest MAPE value is an effective method. The α value used in the double exponential smoothing method from 0.1 to 0.9 is chosen to have the lowest MAPE level: α = 0.6 and α = 0.7. Meanwhile, the weighted moving average method uses three months and four months.Findings: In the MAPE calculation, the results obtained for double exponential smoothing α = 0.6 are 6.45%, and α = 0.7 are 7.14%, while the moving average n = 3 is 13.59%, and n = 4 is 13, 48 %.Implication:It can be concluded that the lowest MAPE value was produced using the DES (Double Exponential Smoothing) method with ???? = 0.6 with a MAPE value of 6.45%. This value is lower than DES ???? = 0.7, 3-month WMA, and 4-month WMA with weights of 0.4, 0.3, 0.2, 0.1.