A picture is worth a thousand words: Translating product reviews into a product positioning map

Product reviews are becoming ubiquitous on the Web, representing a rich source of consumer information on a wide range of product categories (e.g., wines and hotels). Importantly, a product review reflects not only the perception and preference for a product, but also the acuity, bias, and writing style of the reviewer. This reviewer aspect has been overlooked in past studies that have drawn inferences about brands from online product reviews. Our framework combines ontology learning-based text mining and psychometric techniques to translate online product reviews into a product positioning map, while accounting for the idiosyncratic responses and writing styles of individual reviewers or a manageable number of reviewer groups (i.e., meta-reviewers). Our empirical illustrations using wine and hotel reviews demonstrate that a product review reveals information about the reviewer (for the wine example with a small number of expert reviewers) or the meta-reviewer (for the hotel example with an enormous number of reviewers reduced to a manageable number of meta-reviewers), as well as about the product under review. From a managerial perspective, product managers can use our framework focusing on meta-reviewers (e.g., traveler types and hotel reservation websites in our hotel example) as a way to obtain insights into their consumer segmentation strategy.