It's Not You, It's Me: Identity, Self-Verification, and Amazon Reviews

Online retailers often incorporate crowdsourced product reviews to make customers feel more informed and comfortable with online purchases, and thus increase profits. The evaluation of these reviews is also crowdsourced, ostensibly to identify “helpful” reviews. The resulting helpfulness ratings are frequently used as measures for discerning what makes reviews helpful, and are used to determine which reviews are given priority viewing on the site. However, there is no empirical evidence that helpfulness voting reflects customers’ attempts to evaluate product reviews objectively. This study examines review helpfulness voting from the position of the subjective customer rather than the objective anatomy of the review. We develop and empirically test a model, informed by self-verification theory, which explains relationships between online reviewers’ overall opinions of products under consideration (star ratings), product type, and perceived helpfulness of online product reviews. Results suggest that customers’ unconscious attempts to confirm what they already know and believe about themselves, referred to as self-verification, influences helpfulness voting. This work contributes to theoretical understanding of the role of reviews from the users’ perspective and how, through suggesting new ways to identify helpful reviews, human behaviors can inform design of recommender systems.