The Application of the Automatic Interaction Detector (A.I.D.) Programme and Multiple Regression Techniques to the Assessment of Store Performance and Site Selection

Abstract

The paper begins with an examination of the gravity model approach to site selection and concludes that there are certain practical limitations which make it unsuitable for predicting the distribution of consumers' expenditure amongst individual outlets.

It then describes how a combination of the Automatic Interaction Detector (A.I.D.) and multiple regression techniques can be used to provide criteria for judging the performance of a chain of retail outlets and assessing the potential of new sites. A feature of the approach is the measurement of approximately fifty factors which may influence turnover and the inclusion of attitudinal data, derived from a sample survey of shoppers, quantifying factors such as the appearance and image of a store. The method has been successfully applied to supermarkets, durable goods outlets, off-licences, chemists, building societies and car washes.