FORECASTING SALES DATA ON E-COMMERCE USING SINGLE EXPONENTIAL SMOOTHING METHODS

Abstract

Forecasting is predicting the occurrence of something in the future by referring to historical data that occurred in the past that can be used to conduct business analysis. Fairez Shop is an MSME engaged in selling clothing which in its operations generates tens of thousands of data that has never been used for business purposes. Fairez shop also requires a forecasting analysis to predict sales of its products in order to manage supply and help make decisions related to future business. From the problems described above, this research will forecast sales data at the Fairez Shop. The ETL process for designing the OLAP database in this study uses Pentaho Data Integration and the forecasting method used is single exponential smoothing and standard error mean absolute percentage error with R Studio tools, then forecasting results are visualized using Power BI tools. From the trial results, the standard error was obtained in January 2023 of 36.99655%, in February 2023 of 2.817564% and in March 2023 of 2.884921%. With a standard error percentage value below 50%, it can be concluded that this forecasting is feasible and applicable.