Empirical research in marketing should focus on the development of empirical generalizations. Marketers do a huge amount of empirical research, but have little in the way of empirical generalizations. This is primarily because most empirical research consists of ‘original’ or ‘novel’ works looking for significant differences, rather than significant sameness, in unrelated data sets, thus exemplifying the ‘cult of the isolated study’. As a result, the marketing literature is made up largely of uncorroborated, fragmented, ‘one-off’ results. Such results are of little use to marketing practitioners or academicians. We discuss a number of impediments to the development of empirical generalizations – preoccupation with the hypotheticodeductive conception of science, preoccupation with ‘statistical’ rather than ‘empirical’ generalization, the ‘publish or perish’ syndrome in academia, and denigration of replication-with-extension research. We conclude that replication-with-extension research must be championed as the vehicle for discovering empirical generalizations.