Suspicious online product reviews and brand and product characteristics: An empirical analysis with Amazon review data
(A2022-107220)

AUTHORS

Eunhee Ko, Northwestern University; Douglas Bowman, Emory University

ABSTRACT

A suspicious online review is one that could raise suspicion of having a manipulative purpose. To date, these have been studied in the context of services (hotels, restaurants) but little in the context of products and brands. The present paper focuses on products, linking brand advertising and other product-level characteristics to an incidence of manipulative reviews. The authors tested a set of hypotheses using observational data from Amazon.com, arguing that sellers of weaker brands have an incentive for review manipulation. Through numerous analyses and robustness checks, they found that a lower advertising effort was associated with suspicious reviews that were promotional and positive in nature. Also, the authors found that price, product type (tech vs. non-tech), and review sentiment were strongly associated with a suspicion of a manipulative review. The findings imply the importance of the simultaneous consideration of multiple factors when predicting review manipulations.