Forecasting practices: Empirical evidence and a framework for research

Demand forecasting is a relevant issue both in research and practice. In the past several papers have investigated forecasting practices. This research uses data from the GMRG survey to achieve three main objectives. First of all, we aim to describe current practices in the machinery and textile sectors; in particular we investigate: (i) aims and usage of the demand forecast for decision making; (ii) structure of the forecasting process; (iii) algorithms and tools adopted; (iv) both upstream and downstream cooperative forecasting and (v) performance. On a second perspective we investigate contingent variables such as structure (company size and sector), strategy (improvement priorities) and demand characteristics (e.g. number of products) and their relationship with forecasting practices. Finally, the impact of the current forecasting process on both forecasting accuracy and overall company's performance is investigated. In the end we highlight gaps between current research and actual companies' practices; such gaps are discussed to identify areas where support, new tools and concepts are needed to improve companies' practices. © 2007 Elsevier B.V. All rights reserved.