Modeling Preference Change through Brand Satiation

In this study, we develop structural models of preference change due to consumer state dependence through satiation by purchase experience. A dynamic factor model with switching structure is proposed to explain consumer preference changes. Two types of dynamic factor models are separately applied to baseline and satiation parameters in a direct utility model that accommodates multiple discreteness data. The first dynamic factor model has a switching structure for consumer preference, and decomposes brand baselines into time-invariant factor loadings for the coordinates of brand positions and time-varying factor scores for consumer preference directions. The second dynamic factor model applied to satiation parameters extracts the consumer level of satiation in a product category, and this is used as a causal variable in a switching equation to show when and how preferences change over time according to the level of brand satiation. The brand positions and temporal changes of heterogeneous preferences are jointly depicted in a dynamic joint space map. The empirical analysis of a panel dataset shows that our proposed dynamic model, implying that consumers change their preferences when previous brand satiation exceeds the admissible level and preference directions are determined by the previous level of satiation, performs better than alternative specifications, such as a static model with no preference change and a dynamic model without structures which imply that preference changes whenever a consumer purchases a product.