Brands versus brand extensions : implications of brand-based learning for the dynamics of sales

A key precept of brand extension theory is that brand extension, as opposed to new brand, reduces the risk of new product trial in experience goods categories. In this paper we argue that this property has specific testable implications for the dynamics of sales. In particular, in the movie industry—where brand extension takes the form of sequels—sequels' sales ought to be more “front loaded” than originals' sales. On three measures of front-loadedness—ratio of opening week (weekend) box-office to total box-office, average purchase time, and time to peak sales—we verify this prediction robustly. While confirming that new brand extensions are indeed less risky to consumers than new brands, our results also suggest that signaling theories such as those of Nelson (1974) do not have much bite in movies