Article ID: | iaor20116844 |
Volume: | 24 |
Issue: | 3 |
Start Page Number: | 159 |
End Page Number: | 167 |
Publication Date: | Sep 2011 |
Journal: | OR Insight |
Authors: | Fildes Robert, Goodwin Paul |
Keywords: | time series & forecasting methods, demand |
The details of a study of demand forecasting in four supply chain‐based companies are reported. We found that statistical forecasts obtained from expensive software are frequently adjusted using management judgment. Most of these adjustments were small and these often reduced accuracy and wasted management time. However, large adjustments, made for important reasons, tended to improve accuracy. There was also a tendency for the adjustments to lead to over‐optimistic forecasts of demand.