Responses to negative word of mouth: the influence of information overload on purchase intention

Abstract

Purpose The influence of online reviews on consumer behavior is thoroughly documented. Research indicates that negative reviews influence purchase intention more than positive reviews, however, brand holders and retailers have little control over the content of negative reviews. Since many review platforms allow brand holders to respond to negative reviews, potentially mitigating the deleterious effects of the review on the attitudes and intentions of potential consumers. This makes crafting an effective response important, however, researchers are just starting to study the elements of review responses that influence consumer outcomes. This study aims to examine the review, response and consumer factors that influence consumer outcomes. Design/methodology/approach Building on the electronic word of mouth (eWOM) paradigm and using a simulation-based experiment, this study models relationships between a consumer level factor (uncertainty avoidance), review level factor (service failure severity) and a response level factor (information load) and purchase intention. Participants are asked to read a review, a response and then respond to several questions. The model is tested by applying an ANCOVA to data generated from a sample of 887 (approx. 150 per group) participants. Findings Results support relationships between a consumer level factor (uncertainty avoidance), review level factor (service failure severity) and a response level factor (information load) and purchase intention. An interaction between uncertainty avoidance and service failure severity was not supported. Interestingly, an inverted U-shaped relationship between response information load and purchase intention was identified, where less information led to low levels of purchase intention, moderate information led to improved purchase intention and high levels of information led to low levels of purchase intention. Research limitations/implications This study highlights the importance of understanding the elements of responses to negative reviews which influence consumer behavior, and provides evidence that response level characteristics do influence purchase intention. The interplay between review characteristics and response characteristics is still poorly understood (e.g. if the consumer suspects that a review is fake, do they even consider the response in their purchase intention decision?). The influence of review level information load has been tied to purchase intention in previous studies, is there a cumulative effect of information overload between the review and the response? Many research questions remain unexplored. Practical implications These findings carry implications for platform developers and brand managers seeking to mitigate the effects of negative reviews. Specifically, our findings of a u-shaped relationship between response information load and purchase intention indicates that responses of moderate length will be most effective. The finding that when service failure is severe, information load is less effective at mitigating deleterious consumer outcomes indicates that brand holder should prioritize responding to moderate and low service failure reviews. For review platform developers motivated to increase sales, prioritization algorithms should prioritize negative reviews that have a moderate length response over very long or very short responses. Originality/value While online review research is extensive, brand holders have little control over the content of online reviews, limiting the impact of eWOM research. One of the few factors that brand holders can control is the responses to negative reviews, thus studying the response level characteristics that influence consumer behavior promises to impact practice. This is a very recent and very sparce area of exploration. This study contributes to the early steps of this area by applying existing eWOM theory to brand holder responses to negative eWOM. Investigations of the influence of information load of responses to negative reviews have not yet been published.