Applying Diffusion Models with Regional Heterogeneity

Abstract

Recent studies of innovation diffusion have investigated cross-country heterogeneity, but implicitly assumed within-country homogeneity. As such, these studies potentially overlook within-country variations in diffusion patterns, which may be even more important to marketing managers and researchers alike. The current paper is concerned with such intra-country variations using one of many possible a priori segmentation schemes, namely geographic segmentation. It empirically demonstrates that when substantial regional variations in diffusion patterns occur, taking account of these regional differences improves both short- and long-term forecasting under certain conditions. Regional differences in diffusion patterns also provide some important normative implications.