Peramalan Jumlah Transaksi Bulanan pada Agen BRILink Menggunakan Metode Weighted Moving Average (WMA)

Abstract

The rapid development of digital banking services has increased the need for more effective operational planning, including for BRILink agents. One approach to support operational planning is by utilizing historical transaction data to forecast future transaction volumes. This study aims to apply the Weighted Moving Average (WMA) method to forecast monthly transactions at a BRILink agent based on historical data collected over a twelve-month period, from June 2025 to May 2026. The research employed a quantitative approach using secondary data in the form of monthly transaction records, which were analyzed using QM for Windows. The WMA model was implemented with three forecasting periods and weights of 0.5, 0.3, and 0.2, giving greater emphasis to the most recent transaction data. The forecasting results estimated 1,121 transactions for the subsequent period. Model evaluation produced a Mean Absolute Deviation (MAD) of 261.878, a Mean Squared Error (MSE) of 130,054.1, and a Mean Absolute Percentage Error (MAPE) of 18.707%, indicating good forecasting performance. These findings demonstrate that the WMA method provides sufficiently accurate transaction forecasts and can support operational planning for BRILink agents, including cash management, service readiness, and more effective decision-making.