Making the Cut: Modeling and Analyzing Choice Set Restriction in Scanner Panel Data

Abstract

The authors develop an approach to determine and analyze choice set restriction on the basis of secondary source information on consumer purchase histories. Individual-level choice sets are estimated using a Bayesian updating procedure in conjunction with the multinomial logit model. The authors apply the procedure to scanner panel data for the liquid laundry detergent category. An analysis of estimated choice sets across panelists reveals that market share does not “go hand-in-hand” with choice set share (the percentage of choice sets in which a brand is a member). Examining choice set membership patterns, such as the cooccurrence of brands in the same product line, also provides insight into sister-brand cannibalization. Estimation results also show that promotions can expand choice sets, providing excluded brands a means to gain entry and long-term sales benefits.