Thematic Discrepancy Analysis: A Method to Gain Insights into Lurkers and Test for Non-Response Bias

Word of mouth (WOM), long recognized as a highly influential source of information, has taken on new importance with the proliferation of online WOM. Research in online environments has focused on individuals who actively participate in generating WOM. However, over 90% of those that read WOM are non-participants, commonly called “lurkers.” This paper develops and tests a thematic discrepancy analysis (TDA) approach that combines commonly available information on Views and Replies with content analysis to provide new insights into differences between WOM participants and lurkers. TDA provides managers with market-sensing information to identify hidden opportunities and threats, as well as to test for non-response bias. Given the lack of approaches to address non-response bias due to lurkers, TDA represents a significant contribution to research methodology. We demonstrate the efficacy of TDA by applying it to a large scale WOM dataset containing over 80,000 messages from a brand-specific online forum. •Over 90% of those that read word of mouth are non-participants, or “lurkers.”•If lurkers differ from participants, conclusions may be compromised by non-response bias.•We develop a thematic discrepancy analysis (TDA) method to examine differences between participants and lurkers.•TDA provides a means to test for non-response bias in online WOM research.•We demonstrate the efficacy of TDA by applying it to a large scale WOM dataset.