Sales Forecasting: Initial Steps toward the Development of a Computerized Information System for the Integration of Judgmental Forecasts and Computationally-based Approaches

摘要

The need for accurate sales forecasting is increasing, as management attempts to rationalize its decisions. The computationally-based methods that have been developed during the last few decades for mining historical data and producing forecasts are numerous and powerful. But, because of the high degree of complexity of the sales process, there is unfortunately no indication that the sales forecasting issue will be completely resolved by the implementation of these approaches, and that the underlying sales forecasting processes can be fully automated within the enterprise. Hence the necessity of judgmental forecasts which not only complement the computationally-based methods, but allow the development of the highly needed collaborative forecasting processes. In this article, we propose an approach to integrate the computationally-based methods and the judgmental forecasts in order to retain the advantages of both forecasting techniques, while attempting to reduce the difficulties caused by their weaknesses.