This study analyzes the causes of overstocking and implements a more accurate forecasting method for planning the inventory of single-phase prepaid kWh meters at PT PLN (Persero) UP3 Surakarta. The primary issue is the discrepancy between demand forecasting and actual purchases, which is attributed to ineffective forecasting methods and a lack of adaptation to dynamic economic conditions. To overcome this problem, this study uses two methods, namely interviews and ARIMA forecasting. The interview results, which are summarized and visualized in a fishbone diagram, identify the main root causes of the problem as two factors: human and method. From the human side, the lack of communication and coordination between staff and departments is a crucial factor. Meanwhile, from the method side, the use of ineffective forecasting methods and the lack of optimal utilization of historical data resulted in inaccuracies in demand planning. To overcome the method problem, this study applied the ARIMA method to historical data. The analysis results show that the ARIMA (1,0,0) model is the most accurate forecasting model, proven to be significant because it produces a p-value of 0.001 with an error rate below 10% and produces the lowest AIC value. This high accuracy indicates that the model can predict future demand very well. This recommendation provides a significant contribution to PT PLN (Persero) UP3 Surakarta in making more accurate procurement decisions, optimizing inventory management, and minimizing ongoing financial losses.
                        
                        
                        
                        
                            
                                Copyrights © 2025