When chatbots fail: exploring customer responsibility attributions of service failures in disharmonious co-creation contexts

Abstract

The deployment of AI-powered chatbots is driving significant changes in the way that service is delivered and experienced. Given their highly interactive nature, chatbots are an enabler of co-creation, since they encourage customers’ active participation during service production. However, the introduction of chatbots in the frontline adds complexity to the study of cocreation, and implies that co-creation is context dependent. Extant literature on the context of co-creation tends to focus on harmonious collaborations between parties participating in co-creation. However, widespread customer concerns associated with chatbot use may lead to the creation of disharmonious co-creation contexts. Furthermore, it is common for chatbots to fail due to technical errors or lack of customer understanding. When customer expectations are not met, service failures lead to customer disappointment, and trigger a process of attribution of responsibility for failure. The aim of this thesis is to investigate customer perceptions of disharmonious co-creation contexts, and to understand how such contexts may influence expectations, and responsibility attributions, in cases of service failure. Guided by a pragmatist research philosophy, this thesis adopts a mixed methods approach. The initial study, which was qualitative in nature, demonstrated how customers perceive three distinct co-creation contexts when interacting with chatbots: volitional (harmonious); deceptive and coercive (disharmonious). The qualitative study advanced a research framework and accompanying hypotheses regarding the impact of these co-creation contexts on disconfirmation of expectations and attribution of responsibility. The hypotheses were tested through two experimental research studies. Focusing on the investigation of coercive co-creation contexts, the two studies revealed that coercive settings lead to stronger attributions of responsibility towards the company, when compared to volitional settings. The results also show the important role of disconfirmation of expectations as a mediator in the relationship between coercive co-creation and attribution of responsibility. Different failure types and severity levels were evaluated, and the results showed how these two variables, as moderators, may also affect responsibility attributions. Although the co-creation literature has shown a strong reliance on the notion of co-creation as a harmonious, voluntary collaborative activity, this study challenged this view and contributed to the co-creation literature by demonstrating the possibility of disharmonious contexts of cocreation, which may especially arise in human-to-non-human interactions. The results show that despite all the benefits associated with chatbots, it is important to understand why customers may feel forced, or deceived, into interacting with a chatbot, as well as the resulting customer evaluations following such interactions. The findings also suggest a number of strategic implications for managers who are considering the partial or complete replacement of human staff with chatbots in customer service settings.