Can Online Word-of-Mouth Communication Reveal True Product Quality? Experimental Insights, Econometric Results, and Analytical Modeling

Abstract

Online word-of-mouth (WOM) communication in the form of online product reviews, such as those provided by consumers that have previously purchased a product, has become a major information source for consumers and marketers about a product’s quality. The literature has thus used online product reviews to predict a product’s sales and future success, assuming that the average (mean) score of these reviews represents true product quality. However, using experimental and econometric data, this study empirically shows that single point estimation (e.g., the mean) may not be a sufficient predictor of true product quality due to under-reporting by consumers with moderate product reviews.To overcome this concern, this paper first tests the underlying distribution of online product reviews with secondary data from Amazon (www.amazon.com). The econometric results reveal that the reviews for the majority of the products have an asymmetric bimodal distribution. For these products, the mean of the online product reviews does not necessarily reveal the product’s true quality, resulting in misleading conclusions about the product’s future success. In contrast, results from a controlled experiment where all respondents reviewed a product show that their product reviews follow a unimodal (approximately normal) distribution. This confirms that consumers with extreme (positive or negative) views are more likely to review products than consumers with moderate views, resulting in a truncated sample.To identify when single point estimation can be used as a sufficient predictor considering the consumers’ incentive of leaving reviews, this paper develops an analytical model to derive the conditions of when the mean can serve as a valid proxy of a product’s true quality. Finally, we introduce a dual point estimation model using the lower and upper ends of the modal interval of the DIP test to predict true product quality, which is shown to be a superior predictor of future product sales versus the simple and weighted average. The paper concludes by discussing the study’s theoretical and managerial implications.