Collaborative Forecasting in the Fast Moving Consumer Goods Chain
01 / 2011 - 01 / 2015
Forecasting is vital for the activities in a supply chain due to uncertainty and lead- times. Errors in a forecast can propagate upstream, distorting the basis on which decisions are made. Demand variability generally increases upwards the chain the bullwhip e ect. Information sharing and collaborative forecasting can lead to substantial savings, mainly in terms of inventory costs. These bene ts have not been examined in a context of demand uncertainty, and especially forecast model uncertainty, so that the implications for practice are unclear. Promotions, widely prevalent in the fast-moving consumer goods chain, are especially characterised by high demand uncertainty. Forecasting capabilities, in terms of forecast model for- mulation and estimation, are often insu cient at companies, but become crucial when demand uncertainty is explicitly considered. Judgmental forecasting is the dominant form of forecasting in companies, and much of the information that can become available for forecasting is tacit, so that either a judgmental component has a vital role or the tacit information has to be codi ed and captured in a model. As leadtime reduction leads to much larger bene ts than information sharing, the logistic service providers (LSPs) also become important. The value of these two collaborative forecasting approaches, one focused on the manufacturer to reduce inventory costs and one on the LSPs to reduce the delivery lead time, can then be compared, thus acknowledging the wider system context and information-rich context in which supply chains operate and can use collaborative forecasting.