The Divorce of Word and Deed—A Data-Mining Approach to Identify and Evaluate Customer Requirements

The advancement of information technology has made it possible to understand customer requirements through publicly available information, such as online word of mouth. However, using such information alone to guide product design can be unreliable if consumers’ actual purchase actions (deeds) are not taken into account. We propose a framework to understand customer requirements by incorporating consumers’ purchase records as an additional information source and examine the relationship between customers’ expressed satisfaction (words) and their actual purchase (deeds). Utilizing unique data sets in a cell phone market, we mine and characterize each product feature in a two-dimensional model based on how important they are to customers’ expressed satisfaction as well as to the actual purchase, respectively. We find that, consumers’ words and deeds are not always consistent: whereas some requirements are expressed as important features in determining customer satisfaction in words, they do not stand out in factors that influence consumers’ purchases; there are also requirements that are silent in words, whereas customers value them when making a purchase. These findings highlight the importance of incorporating the deed dimension into customer requirements analysis. By comparing product requirements across these two dimensions, our framework offers actionable guidelines for product design, including which features are worth investment for improvement/innovation and which features may be overemphasized in review-based analyses.

History: Bin Gu, Senior Editor; Yili (Kevin) Hong, Associate Editor.

Funding: J. Feng acknowledges funding by the National Natural Science Foundation of China [Grants 72442015, 72171132], and X. Sun acknowledges funding by the NSFC [Grant 72471183].

Supplemental Material: The online appendix is available at https://doi.org/10.1287/isre.2022.0433.