MODEL PERAMALAN PENJUALAN MENGGUNAKAN PENDEKATAN WEIGHTED MOVING AVERAGE PADA UMKM KOPI

Abstract

Sales forecasting plays a critical role in supporting effective inventory management, particularly for small and medium enterprises (SMEs) in the coffee sector, where demand exhibits fluctuating patterns over time. Inaccurate demand estimation may lead to stock shortages, resulting in lost sales opportunities, or overstocking, which increases holding costs and product waste. Therefore, a reliable and accurate forecasting model is required to support optimal inventory planning and operational efficiency. This study aims to develop and implement a sales forecasting model using the Weighted Moving Average (WMA) method and to evaluate its predictive accuracy. The WMA method, as a time series approach, assigns different weights to historical data, emphasizing more recent observations to better capture short-term trends. Historical coffee sales data were utilized as the basis for model development. Forecast accuracy was evaluated using the Mean Absolute Percentage Error (MAPE). The results indicate that the WMA model provides a high level of accuracy, with a forecast value of approximately 22,914 units for the subsequent period and a MAPE value of 0.29% using weight parameters of 0.7, 0.2, and 0.1. This level of accuracy is categorized as very high, demonstrating that the proposed model is effective and reliable for supporting inventory decision-making and improving operational performance in SMEs.