How product review voting is influenced by existing votes, consumer involvement, review valence, and review diagnosticity

Websites that allow consumers to post online reviews often allow others to signal the value of the reviews by voting for them as being helpful (i.e., an upvote) or unhelpful (i.e., a downvote). Although a review's content should drive votes, herd behavior may also be operative if consumers discount their own judgments of the review based on signals others provide. Using a field study of online product reviews from Amazon, a longitudinal experiment, and two cross-sectional experiments, we find evidence for herd behavior: consumers' voting decisions and intentions are influenced by others' votes. However, this effect is complex with multiple moderating factors, including consumer involvement, review valence, and review diagnosticity. We discuss the theoretical and practical implications of these findings for reducing the potential negative effects of herd behavior in the context of online reviews.