Reaching for the stars: consumers’ interpretations of online rating distributions and their validity as an indicator of product quality

Abstract

Over the past two decades, rapid advances in technology and the omnipresence of the Internet have led to a fundamental change in our shopping behavior. While purchase behavior in traditional bricks-and-mortar stores is constricted by, for instance, limited retail spaces and finite opening hours, the Internet enables customers to shop anything, anytime, and anywhere. Moreover, while in the past, consumers were reliant on the quality of sales people’s advice or recommendations from their friends, they can now share their experiences and opinions about products, services, companies, and brands on a variety of websites such as Amazon, TripAdvisor, and Google with anyone. Investigating consumer reviews on an aggregate level, a great deal of literature is centered toward consumers’ response to different characteristics describing the distribution of rating scores, including, for instance, average product ratings, the dispersion of rating scores, and rating volume. However, despite the abundance of research on consumers’ reactions to different characteristics of rating distributions our knowledge of the effects caused by further distribution features (e.g., mode, median, skewness, etc.) is still limited. Adding to this stream of literature, the first empirical research paper presented in this thesis demonstrates a tendency to use the mode as a heuristic basis when making product inferences from online rating distributions in such a way that product evaluations inferred from rating distributions with an equal average, standard deviation, and number of ratings systematically vary by the location of the mode; a phenomenon referred to as the mode heuristic. The results of a series of six studies, using a mix of experimental and real-world data, (1) provide strong empirical evidence for the existence of the mode heuristic in a variety of different contexts, (2) shed light on this phenomenon at the process level, and (3) demonstrate how product inferences based on the mode heuristic depend on the visual salience of the mode. Furthermore, several researchers have questioned whether the proliferation of online consumer reviews should be considered a positive development from a consumer welfare perspective by investigating if the evaluations posted online can actually reflect the ‘true’ quality of a product. In this vein, de Langhe et al. (2016a) reported a substantial gap between the extent to which consumers trust in average ratings when making inferences about the quality of a product and the actual validity of such ratings as an indicator of a product’s ‘objective’ performance. The second research paper presented in this doctoral dissertation replicates and extends this finding by examining how the convergence between objective and rated quality alters over the product life cycle and investigating which quality indicator is a better predictor of sales performance. Overall, this doctoral dissertation contributes to a better understanding of consumers’ use of online reviews in purchase decision making and offers highly relevant implications for theory and practice.