Targeting Mr. or Mrs. Smith: Modeling and Leveraging Intrahousehold Heterogeneity in Brand Choice Behavior

Abstract

We develop a hierarchical choice model to account for the choice utility heterogeneity of individual shoppers that belong to the same household. Our model allows us to measure how much variability in purchase behavior exists among individuals in a household, and to compare this to the variability that exists across households. Because of the presence of multiple shoppers from the same household, we also extend the concept of household-level state dependence to consider state dependence at the individual level. We apply our model to five different grocery categories. We find that the intrahousehold heterogeneity in estimated brand intercepts and (to a lesser extent) price sensitivities is about 20%–30% of the interhousehold heterogeneity in these parameters. However, with promotion sensitivities, we find intrahousehold heterogeneity, in most cases, to be as large as interhousehold heterogeneity. Our state dependence results show that past brands purchased by an individual have a much stronger influence on subsequent purchases than those purchased by anyone in the household. We use our estimated utility parameters to compare the expected profitability of promotions targeted at the individual rather than at the household and find substantial (more than 50%) improvements in the incremental revenue of supermarket promotions. Data and the online appendix are available at https://doi.org/10.1287/mksc.2018.1088 .

Tightwads and spendthrifts: An interdisciplinary review

“…More broadly, the replicability of all findings reported in this review should be investigated. 3 In a study of how different household members respond to promotional discounts, Bruno, Cebollada, and Chintagunta (2018) reported conceptually similar results. In particular, they found that, in terms of sensitivity to price promotions, "two individuals living in a household are expected to be as similar as two individuals picked at random from the population" (p. 17).…”

A novel response model and target selection method with applications to marketing

“…Moreover, an aggregate advertising response model based on consumer population dynamics was proposed by Wang et al (2013), while Li & Ansari (2014) proposed a Bayesian semi-parametric approach for endogeneity and heterogeneity in choice models. Bruno, Cebollada & Chintagunta (2018) developed a new model and used it to deal with the intra-household heterogeneity in customer brand choice behaviour. Kappe, Blank & DeSarbo (2018) developed a random coefficients mixture hidden Markov model for flexible patterns of unobserved heterogeneity in both the state-dependent and transition parameters.…”