The Effect of Expressing and Receiving Negative Word-of-Mouth on Purchase Behavior
Social media have enabledconsumers to express their thoughts and feelings openly after consuming a product or service. In particular, customers who have a dissatisfying experience can post their negative reactions to an online public forum such as a company’s website, social networking sites, or third-party complaint websites. This negative word-of-mouth (NWOM) is observable not only by other customers, but also by the firm itself. Thus, these social technologies motivate a firm to understand the sentiments and the consequences of these negative messages from its consumers.
Previous work on NWOM and consumer complaint behavior (CCB) found that consumer complaintsaretriggered by negative feelings or emotions from perceived dissatisfaction, and such emotions are believed to affect attitudes towards brands, purchase intentions, and purchase decisions. However, previous NWOM researchis limited as follows: (1) WOM valence was simply coded as positive, neutral, or negative, but did not look at different types of emotions; (2) the effect of expressing and receiving NWOM was not distinguished although these activities may involve different cognitive or emotional mechanisms; and (3) most studies were not able to estimate the impact of NWOM on actual purchase behaviors.
This paper addresses these shortcomings by conducting a sentiment analysis on customers’ NWOM messages and linking the NWOM activities to actual purchase records. Based on a review of previous literature on the effect of posting and viewing NWOM, we pose two separate hypotheses on the effects of expressing and receiving NWOM on purchase behaviors, as well as two research questions on different types of negative emotions in NWOM messages and their differential effects on purchase behaviors.
Our data come from a large coalition loyalty program that maintains a social media forum. Members earn miles for purchases at sponsors in different product categoriesandredeem earned miles for gift cards, merchandise, airline tickets, etc.At the end of 2011, the loyalty program announced a policy change that affected all existing customers and triggered NWOM on their online community, creating a natural experiment. The company provided us with transaction and redemption records of posters, viewers, and a control group, as well as their records of posting activities and viewing logs. To test the effects of expressing and receiving NWOM, we estimate a random intercept model. We identify the types of emotions embedded in NWOM and quantify the level of intensity of such emotions with a textual analysis.
We find the interaction effect between posting and redeeming to be positive, and viewing to be negative. Regarding emotions, anger has a negative but insignificant effect, while concern has a positive effect. The results suggest that (1) NWOM receivers become aware of the negative social climate about a brand and the disadvantages they may experience,and then decrease their spending, (2) expressing negative feelings helps consumers resolve their anger to some degree, and when an opportunity to realize the value of a brand is combined, expressing NWOM can have a positive impact on purchase decisions, and (3) emotions play a role in determining the direction of posting effects, thus we need to identify the different emotions expressed in NWOM messages to better estimate their differential effects.
This study has important implications for direct and interactive marketing, which has traditionally focused on managing customer relationships to maximize lifetime value (CLV). CLV measures value to marketing organization, but does not account for value to the consumer. With consumers empowered through social media, organizations should seek to maximize CLV and value to the consumer. The organization chose not to respond to the NWOM. Our results suggest that a better strategy would have been to respond by reminding consumers of the value of the brand, perhaps with an incentive to engage with the brand by redeeming miles. Engagement by verbalizing dissatisfaction and redeeming miles should increase future purchase behaviors.