This study deals with the measurement of the effects of retail marketing instruments on annual sales in retail stores. We assume that the sales level in retail stores is determined by an interplay of supply capacity and demand factors. In some stores sales are supply-determined, whereas in other stores sales are demand-determined. If it is not known a priori what economic regime applies, the more traditional approaches lead to biased estimation results. Therefore, a switching regression model is proposed to estimate the marketing mix effects. Our ideas are tested using data from four different types of stores in the Dutch retail trade and a comparison is made with a more traditional approach. The main conclusions are: the traditional approach leads to underestimation of the marketing mix effects. A switching regression model seems to be a promising instrument for analyzing these effects. The method has a wider applicability than the retail trade.

retail trade,
ERIM Top-Core Articles
International Journal of Research in Marketing
Erasmus Research Institute of Management

Bode, B, Koerts, J, & Thurik, A.R. (1988). On the measurement of retail marketing mix effects in the presence of different economic regimes. International Journal of Research in Marketing, 107–123. doi:10.1016/0167-8116(88)90063-8