Application of moving average and single exponential smoothing (SES) methods for frozen food sales forecasting at Rezeki Baru Store

Muhammad Hafizh Syah Utama (Universitas Muhammadiyah Sumatera Utara)
Zainal Azis (Universitas Muhammadiyah Sumatera Utara)

Jurnal Sains, Teknologi dan Komputer
Vol. 3 No. 2 (2026): Jurnal Sains, Teknologi & Komputer (SAINTEK)

This study aims to apply and compare the Moving Average (MA) and Single Exponential Smoothing (SES) methods for forecasting frozen food sales at Toko Rezeki Baru. The study utilizes historical sales data from the 2023–2025 period. Both methods were implemented in a web-based forecasting system, and their accuracy was evaluated using Mean Absolute Deviation (MAD), Mean Squared Error (MSE), and Mean Absolute Percentage Error (MAPE), with MAPE serving as the primary indicator for determining the best method. The results indicate that SES yielded lower error values than MA for the majority of products, demonstrating superior performance in generating accurate forecasts. A comparison of MAPE values across 31 products revealed that SES produced lower error rates for 28 products, whereas MA performed better for 3 products. The developed system can be used to forecast sales for 2026 and assist Toko Rezeki Baru in inventory planning and stock management decision-making.