Content area
Full Text
Introduction
Recent advancements in artificial intelligence encourage more and more firms to use chatbots for service delivery and incorporate them into the frontline (van Doorn et al., 2017). Chatbots are text-based virtual robots that emulate human-to-human conversation through natural language processing (Schuetzler et al., 2018; Wirtz et al., 2018). They offer the chance to provide efficient customer service around the clock, therefore serving as a crucial strategic asset for firms (Thomaz et al., 2020). For instance, a recent industry report forecasts that by 2025, 95% of all consumer interactions with a firm will be powered – that is, augmented or replaced – by chatbots (Servion, 2020).
In contrast to traditional self-service technologies with a merely functional character, chatbots are equipped with additional social–emotional and relational elements (Wirtz et al., 2018). Not only does the natural language interface remind of human conversation (Tuzovic and Paluch, 2018), chatbots also take on roles that were so far fulfilled by human frontline employees and provide personalized responses based on sophisticated speech recognition tools that create an anthropomorphic conversation (Nass and Moon, 2000; Wilson et al., 2017; Wuenderlich and Paluch, 2017). However, this rapid advancement of chatbot technology comes with a dark side: As chatbots become increasingly anthropomorphic, consumers find it increasingly challenging to correctly distinguish between human or artificial conversational partners (Candello et al., 2017).
As this challenge gains traction, firms are confronted with the question whether or not to disclose information on the nonhuman identity of chatbots. Previous studies that attempt to address this question consistently find negative consumer reactions (i.e. perceiving it as less empathetic or knowledgeable; Luo et al., 2019) to disclosed vs undisclosed chatbots and therefore stress the detrimental effects of chatbot disclosure. As these negative reactions in service interactions jeopardize customer retention by alienating customers (Puntoni et al., 2021), this study aims to answer the following research question:
How does disclosing chatbot identity influence customer retention?
In order to find ways for mitigating or at best reversing potential negative retention effects of chatbot disclosure, firms need to understand the mechanism that fuels the relationship between chatbot disclosure and retention. Approaches that try to explain negative reactions highlight consumers' aversion toward algorithms that is rooted...