FY Company’s Product Demand Forecasting Research

FY watch has achieved its fullest possibilities alongwith the economic progress and Internet popularity. At the same time, the uncertainty of market demand intensified by the complexity and diversity of distribution channels also presents great challenge to its demand management. With a 52% shutdown rate, 233 out of 449 FY’s existing watches are no longer in production. Even worse, the bestsellers fail to meet the market demand,which result in a 32% of dead stock. Thus, in order to improve customer satisfaction and raise the brand value of FY, it is urgent to establish the collaborative demand forecasting system of FY by the study of the scientificity and accuracy of product demand forecasting.My dissertation firstly analyses the current status of FY’s demand management from the perspective of product classification, the accuracy of demand forecasting rate and the control of dead stock. The analysis reveals that FY has made several missteps over the accuracy of product demand forecasting, the sufficiency in product classification forecasting, the consciousness of management on the upstream and downstream of supply chain and the perfection on collaborative forecasting. In view of these problems, my dissertation proposes a new FY demand forecasting system under supply chain. The system sorts watches into ABC-XYZ classes by the natures of products and choose different forecast models under different classes. My dissertation takes an AX watch C.LA8262.GWRD of clover series as an example to elaborate the build of a quantitative forecast model. The first step is to build assembled forecast models, which are combinations of ARIMA model, Winter exponential smoothing model, grey system and BP neural networks. FY then chooses the best forecast model and evaluate it by error prediction. In this way, FY can optimize its demand management and increase the accuracy of demand prediction.