The authors determine the multiple effects of retail promotions on brand loyal and brand switching segments of consumers. Segments are determined by an iterative Bayesian procedure. The variations in within-segment brand shares within a store are related to promotional variables by a logit model estimated by nonlinear least squares. Store share is modeled as a function of store attractiveness, a summary measure of the store’s promotional activity on the multiple brands. Finally, category volume is related to overall product category attractiveness in a model that includes bath current and logged effects. The research approach is applied to the IRI ground coffee data. Results include: (1) the market can be characterized by brand loyal segments, each of which buys mostly their favorite brand, and switching segments, each of which switches mainly among different brands of the same type (e.g., drip, percolator), (2) promotional variables have significant effects on within-segment market shares, the effects being different across segments, (3) store share is related significantly to promotional attractiveness of a store, (4) the overall promotional attractiveness of the product category has significant current and lagged effects on category volume, and (5) the logged effects resulting from consumer purchase acceleration and stock up last longer for brand loyal segments than for switching segments.
Â