Conversational Agents in Service Context: Towards a Classification of Human-like Design Expectations

The increasing application of Conversational Agents (CAs) changes the way customers and businesses interact during a service encounter. For instance, chatbots are the first point of contact for many customers. In this context, prior research has shown that CAs implemented with a human-like design lead to improved service satisfaction, perceived service quality, and trustworthiness, among others. Developers have become accustomed to adopting a one-size-fits-all approach by designing CAs human-like. However, not every human-to-human service requires the same social and human-like interaction (e.g., contract cancelation vs. doctors’ appointment), which has also been shown in current research. At present, existing research lacks a synthesis of the relationship between CA design and service encounter context. Against this background, we conducted a literature review and derive classifications based on the dimensions of service context (professional/private) and human-like design (low/high), which enables the identification of relevant research gaps and related literature.