Can online word-of-mouth communication reveal true product quality?

Online product reviews help consumers infer product quality, and their mean (average) rating is often used as a proxy for product quality. However, two self-selection biases --- acquisition bias (mostly consumers with a favorable predisposition acquire a product and write a product review) and under-reporting bias (consumers with extreme, either positive or negative, ratings are more likely to write reviews than consumers with moderate product ratings) --- render the mean rating a biased estimator of product quality. Integrating empirical data from multiple websites and product categories with a field study that consistently show a positively-skewed, asymmetric, bimodal (“J-shaped”) distribution of online product reviews, we derive that the mean rating of online product reviews does not represent a product’s own quality in absolute terms or relative sense (comparing relative quality across products) due to existence of both acquisition bias and under-reporting bias. To understand the nature and effects of these two self-selection biases, we analytically derive the dynamic process by which acquisition and under-reporting biases shape how consumers make product purchasing and reviewing decisions, and how their decisions shape the distribution of online product reviews over time. This process consists of consumers: a) reading existing online product reviews to form their prior product quality expectations; b) updating the prior product quality expectations with their individual preferences to observable attributes; c) making their purchasing decisions based on their expected utility; d) choosing to write a review based on the difference between their realized utility after experiencing the product relative to the expected utility before purchase; e) any new review, if written, is being added to the pool of online product reviews that affects the quality expectations of subsequent consumers. The relative role of the two self-selection biases in the dynamic process of shaping consumer product quality expectations and affecting purchasing decisions are analytically and empirically examined. Empirical results reveal that consumers do realize the two self-selection biases and attempt to correct for them; however, consumers cannot perfectly account for the proposed self-selection biases because of bounded rationality. Due to bounded rationality, consumer surplus suffers from both self-selection biases. A sales forecasting model with empirical data from Amazon.com shows that consumers use other parameters of the distribution of online product reviews besides the mean rating to infer product quality, supporting the notion that consumers are boundedly-rational who attempt to, but they cannot fully account for, self-selection biases.