An Exploratory Study: 'Blind-Testing' Consumers How They Rate Helpfulness of Online Reviews

摘要

Amazon.com uses the helpfulness vote (H-Vote) on consumer product reviews to signal quality. The top reviews with the best H-Vote are thus prominently displayed so that consumers readily use them for their purchase decision-making. However, the influence of those H-Votes may not be the same depending on contexts. We conducted a pilot survey experiment by using 108 consumers to investigate. The results showed that consumers were not swayed by a positive or negative H-Vote for a top favorable review in Amazon.com. However, consumers significantly voted positively for a top unfavorable (or critical) review in Amazon.com when a negatively-biased H-Vote was shown. To further explore the underlying factors, we deconstructed the YES/NO-based H-Vote with a Likert-scale-based helpfulness rating for (a) product learning and (b) purchase decision-making. While the difference in H-Vote did not affect the learning rating, it did impact the decision rating under a false-negative H-Vote. Implications and future research directions are discussed.