A Brand's Eye View of Response Segmentation in Consumer Choice Behavior

An approach is developed to reveal segmentation in response to marketing variables at a brand-level perspective. In the proposed procedure, response segmentation is analyzed separately for each brand rather than jointly across all brands. This yields a segmentation picture oriented toward the potential targeting objectives of the brand manager. Using the multinomial logit and probabilistic mixture models, the procedure first calibrates consumer response in the brand choice decision. Individual-level measures of response for a given marketing variable (for example, price) are then computed, and brand-level segments are obtained by clustering brand-specific response. Using scanner panel data, the approach is applied to price response segmentation for brands that compete in the ground coffee category. The results illustrate a series of implications brand strategy, particularly the potential for targeting marketing activity to different response segments.