Toward Personalized Online Shopping: Predicting Personality Traits Based on Online Shopping Behavior

Abstract

Consumer's personality traits have a strong influence on their shopping behavior. Hence, e-tailers, rather than merely targeting broad consumer segments, should tailor their shop to those personality traits. However, there is no guidance on how e-tailers can assess a consumer's personality without relying on self-reported data. This study shows how consumers' personality traits can be predicted solely from their online browsing behavior. In a large-scale study, we demonstrate that a machine learning algorithm can predict the personality traits Need for cognition, Need for arousal, Lay rationalism and each of the Big 5 personality traits with accuracy comparable to well-known studies relying on social media data. We also establish that our algorithm is reliable in its predicted probabilities and is capable of making predictions of multiple personality traits in real time. Our research shows that e-tailers can quickly determine a consumer's personality traits and then dynamically adjust their online shop accordingly.