IMPROVING COLLABORATIVE FORECASTING WITH FUZZY DELPHI IN FOOD SUPPLY CHAIN (Case study in coffee shop retailer)

Abstract

Collaborative forecasting is a method that derived from Collaborative Planning Forecasting and Replenishment (CPFR) that has goals on the information sharing between two parties in exchange of necessary information to generate single forecast between two partners. Collaborative forecasting has been widely used in multiple industries. Yet, there is scarcity of the research on the studies of collaborative forecasting in food supply chain industry as already highlighted by multiple researches. Furthermore, the main problem of the collaborative forecasting in food supply chain studies is the scarcity of the studies that examine how the supply chain stages conduct long term and accurate collaborative forecasting. Thus, fellow researcher Eksoz et al. create framework that helped to overcome the problem. Another problem arises when Eksoz et al. framework were not providing specific way to determine the appropriate forecasting technique that should be applied. Based on the issue, researcher has tried to improve the framework by using fuzzy Delphi method to address the forecasting technique and judgmental adjustment problem as proposed by the framework. This research uses fuzzy Delphi method to improve the collaborative forecasting framework to adjust forecasting result based on expert opinion and asses the accuracy improvement. It was found from the case study in coffee shop after implementing the collaborative framework combine with fuzzy Delphi that it needs five levels adjustment that translates Likert-Scale to the fuzzy method to adjust forecast result based on expert opinion. From the result, it is showed that accuracy improvement after the adjustment is 93% and 34% for Mean Squared Error and Tracking Signal, respectively.