Human vs. Chatbot. Who is perceived as more trustworthy?

Abstract

In response to the increasing importance of e-commerce, new communication modes have been developed. For example, chatbot technologies were designed to interact efficiently with customers. As trust can have positive effects on purchasing intentions, trust is the basis of successful business relationships. Trust is also fundamental to the virtual environment and, therefore, users’ trust is a prerequisite of conversational agents, such as chatbots. According to the model of trust, the main judging dimensions in the trust-formation process are warmth and competence. There is still little research on how these trust assessments are made in virtual contact. Moreover, existing research does not provide clear results about the extent of the difference in the interrelationships of sympathy, competence, and trust between human-machine interactions and human-human interactions. Therefore, the present study addresses this gap in research and analysis of trust assessments in human-machine and human-human sales interactions. To test our hypotheses, an experiment was conducted that included a chat-based live sales interaction which the test subjects participated in and evaluated. 98 people took part in the simulation. The results show that the warmth of human-human interaction is perceived as equal to that of human-chatbot interaction. No differences were found in the competence perceptions of the human-chatbot or human-human interactions.