Mind Your Text in Marketing Practice

In this commentary, I address several topics. First, I list a set of typical marketing questions for text analytics. Then I discuss a few common, yet perhaps nonobvious, complications for text analyses. Following this, I suggest an initial path for those who are new to text analytics for marketing. Finally, I illustrate multidimensional sentiment analysis, an extension of sentiment methods described by Berger et al., using shared R code and an online text source. Overall, I urge analysts to incorporate a large degree of inspection and qualitative analysis and to carefully consider the characteristics of both their texts’ origins and any dictionaries used in analyses.

• Data Mining and Modeling

• Economics and Electronic Commerce