In this paper we propose a sequence of steps for setting up and maintaining a forecasting system for predicting future demand of products or services. Our main objective is to propose a methodology that allows forecasting techniques to be used as a supporting tool in managerial decision making. Our procedure is implemented in six main steps, covering issues such as product classification and the analysis of some software available to proceed with the model calculation. We also review the main time series analysis and their application. In short, such models allow the analyst to project into future periods patterns and trends that were recognizable from the analysis of past demand data. A case study from the food industry illustrates the forecasting techniques reviewed and the steps of the method we propose.
forecasting system; product classification; time series analysis