Competitive analysis of online reviews using exploratory text mining

Purpose – This paper explores the usefulness of analyzing text-based online reviews using text mining tools and visual analytics for SWOT Analysis, as applied to the hotel industry. These results can be used to develop competitive actions. Design – The text mining/visualization tool, ReviewMap, was used to transform an archive of reviews spanning multiple suppliers into a hierarchy of data of increasing dimensionality. Visual summaries at each level were integrated to propagate selections at one level throughout the rest of the hierarchy. These visual summaries identify features required for competition at a given level and features that currently discriminate amongst competitors. Methodology – The approach was exploratory, the objective of which was to determine if useable competitive intelligence could be found in a typical collection of online reviews for a set of competing hotels. A publically available collection of reviews was subjected to a set of text mining procedures and visual analyses in order to summarize the features and opinions expressed. Originality – Prior analyses of online reviews relied solely upon numeric “star” ratings. This study utilized text mining to uncover information within the written comments and applied the information in a SWOT Analysis of three competing hotels. Findings – In the set of reviews used in this paper, a common measure of analytical power almost doubled when text mining summaries of the written comments were used in combination with numeric ratings. Visual analytics revealed the dominant features for each hotel, the features required of all hotels competing at a given level, and the features that define specific positions within the competitive landscape. This analysis of strengths, weaknesses, opportunities and threats revealed several promising competitive actions for the hotels in the study.