A Comparative Study of Demand Forecasting for Aftermarket Parts in Heavy Equipment Industry (PT XYZ Case Study)

Abstract

The global economic crisis has reached the world today, forcing many customers to become more cost aware in their search for better quality and service, and forcing corporate organizations to discover more effective and efficient ways to compete among them. The main objective of this research is to choose the best forecasting method to predict the demandfor spare parts at PT. XYZ highly fluctuating, and to avoid or minimize stockouts. The demand for high-priced spare partsand capital goods is considered discontinuous if it is random and contributes a large part of the inventory value. Fluctuating demand for goods will be difficult to predict, and inaccurate estimates can cause huge losses for the company due to obsolescence of spare parts or unfulfilled demand for spare parts. Running a successful company operation today requires organizational strength to supply the needs of its customers. This study discusses the appropriate demand forecasting method for the fluctuation demand for spare parts products at an Indonesian companynamed PT. XYZ. This study compares four forecasting methods to predict the demand for spare parts at PT. XYZ is the ABCAnalysis, the Moving Average, the Simple Exponential Smoothing (SES) and Exponential Smoothing (ES) with TrendAdjustment. This study uses demand data for 2017-2020 to forecast demand in 2021 and uses the optimum alpha value of0.4065 for SES which is obtained by calculating using MS Excel Solver and uses alpha value of 0.5 and beta value 0.3 for ES with Trend Adjustment. The performance of this forecasting method is determined based on the smallest mean absolute percentage error and the level of forecasting accuracy (tracking signal) which is close to zero, and the results of this study indicate that the use of the Exponential Smoothing with trend adjustment method has the best performance compared to the other three methods.