Text analysis of online customer reviews for products in the FCB quadrants: Procedure, outcomes, and implications

Abstract

The present researchers performed a study to present a detailed procedure of full-scale text analyses of online customer reviews (OCRs) and display an exemplary trial of machine-learning-based generation of “selling phrases” from a subset of the same database. As an effort to overcome the limitations of past studies, the present study expanded the scope of target products from one or a few to 16, based on the classic and long-held product classification scheme of the FCB Grid, to obtain broad understanding of the general market. The results are thought to have significant practical values for marketers, consumers, and marketplace platforms. Of particular note, the total of 48 word clouds yielded results that are overall consistent with what the FCB Grid suggests, and the phrases generated through the machine-learning algorithm were considered to have potential to supplement or substitute for the earlier step(s) of the “attributes-consequences-values” chain in the popular Means-End Chain (MEC) model. Also discussed extensively are the immense value of OCR data, as they are more “real,” “natural,” “fair,” and “live” than those acquired from laboratories or through formal surveys or interviews, and the possibility that the data could be utilized for revisiting and reconfirming existing beliefs in marketing research.