Market segmentation revisited: the predictive utility of demographics

This paper compares the predictive power of demographic variables to that of other, more complex segmentation variables. To achieve this objective, the mean R-square of 20 demographic models obtained from a representative survey was judged against mean R-square of 34 non-demographic models drawn from existing published studies. The result indicates that demographics have significantly higher explanatory power than some well-established theories. Thus, our finding strongly refutes the proposition advanced by a number of prominent market segmentation researchers that the relationships between demographic segments and marketing variables are too weak to be of any practical value. Based on the result of this analysis, and given the fact that only five variables were used in the demographic models, it is concluded that demographics are, at least as useful as other segmentation bases.