A Brand Loyalty and Attachment-Based Bayesian Brand Choice Model

Abstract

We investigated how brand attributes, familiarity and consideration, and residuals of prior marketing mix exposures on individuals influence the intention to purchase. We employed a Bayesian view point to make probability statements about the parameters and express uncertainty in a probability or distribution statement. We used a behavioral data set of US auto industry. Nine thousand six hundred seventy-five observations were used in the analysis. Prior literature neglected to take into account the customer-brand relationship variables which may explain unobserved customer heterogeneity. Our brand choice model builds on a set of behavioral variables and focuses on the brand dimensions. Within a Bayesian framework, we proposed a new brand choice model which accounts for the customer-brand relationship. Our mixed model showed familiarity and brand loyalty vary by segment; brand loyalty is the most important predictor of purchase intention for each segment. Besides, our mixed model indicated fixed effect of the interaction between familiarity and attachment to be a significant predictor of purchase intention. We showed that consideration positively and brand attachment negatively are associated with purchase intention. Our results regarding negative effect of brand attachment on purchase intention suggest further investigation of possible moderating factors of the relationship between brand attachment and purchase intention.