Perbandingan metode Double Exponential Smoothing dan Simple Moving Average pada kasus peramalan penjualan

Abstract

Barokah Building Shop (TB. Barokah) located in Singosari, Malang, East Java is one of the shops that sells building materials. Some of the items being sold are white cement and black cement. The number of cement sales every month is very diverse and fluctuating. Black cement occupies a very high number compared to black cement on the sales chart every month. Likewise with the black cement from the Bosowa brand, white cement from the Gresik brand and the Tiga Roda brand, although demand is small, there is still a movement in the number of sales. The fluctuating amount of demand made the amount of product inventory prepared uncertain, which was influenced by the number of types and brands, so that the product inventory management had difficulty in supplying it. This study aims to predict the amount of cement sold in the following month, as well as to determine the performance between the two double exponential smoothing (DES) methods and the simple moving average (SMA) to forecast cement sales results in TB. Barokah. The SMA method is able and can perform forecasts with stable/constant demand or sales data. Meanwhile, the DES method is able and can provide values to the stratified weights with up to date data. DES is capable of forecasting cement sales each month with an average percentange error (PE) of 0.14%, while the SMA is with an average PE value of 1.35%. Based on the test results using data from TB. Barokah found that the most effective method is DES because it has a smaller PE value than the SMA.