Opening the Black Boxes of Consumer Misbehaviors: Insights from Studying Online Trolling

During the past decade, researchers and practitioners have shown an increased interest in how people misbehave in their role as consumers (Daunt and Greer 2015). Several studies have explored how some consumers act "in a thoughtless or abusive way, causing problems for the firms, its employees, and other customers" (Lovelock and Wirtz 2016, 524). Most of these studies have taken a dispositional perspective, explaining consumer misbehaviors by referring to the characteristics and predispositions of misbehaving consumers (Daunt and Greer 2015). At the same time, perpetrators of consumer misbehaviors have been in the foreground of existing misbehaviors-management practices. This paper offers and advocates an alternative approach to understanding and managing consumer misbehaviors. Using actor-network-theory (ANT), we suggest that consumer misbehaviors cannot be attributed only to perpetrators but should rather be seen as effects of networks of all kinds of actors (Latour 2005). To better understand the assemblages that allow and perpetuate consumer misbehavior, we studied online trolling behaviors. Trolling involves deliberate, deceptive, and mischievous attempts to provoke reactions from other online users. In practice, trolling includes anything from offensively replying to customers under fake customer service accounts or posting misleading and damaging product tutorials to communicating with other online users in humorous but also aggressive ways. These behaviors closely correspond to the behaviors of so-called problem customers (Bitner, Booms, and Mohr 1994) and jay customers (Lovelock and Wirtz 2016). Their pervasiveness, their potential impacts on marketers, businesses, brands, and other consumers, and the fact that they are under-researched phenomena (Buckels, Trapnell, and Paulhus 2014) make trolling behaviors a relevant and insightful research context for investigating consumer misbehaviors. In our study, we examined (1) what human and non-human entities are assembled in the performance of trolling and (2) what roles these entities play in the making of this form of online misbehavior. To do so, we adopted a case-study approach, investigating five different instances of trolling: playful trolling, shock trolling, online pranking of other consumers and businesses, fake customerservice trolling, and "good old-fashioned" trolling. The research data were drawn from 250 hours of nonparticipant observation of actors and their practices, seven in-depth int