Enhancing customer satisfaction with chatbots: the influence of anthropomorphic communication styles and anthropomorphised roles

Ying Xu,Jianyu Zhang,Rui Chi,Guangkuan Deng

DOI: https://doi.org/10.1108/nbri-06-2021-0041

2022-02-15

Nankai Business Review International

Abstract:Purpose Chatbots are increasingly used in online retail settings and are becoming a powerful tool for brands to engage customers. However, consumers’ satisfaction with these chatbots is mixed. Thus, this paper aims to investigate how using a social- versus task-oriented anthropomorphic communication style can improve customer satisfaction. Design/methodology/approach The authors explore the link between the anthropomorphic communication style use and customer perceptions/customer satisfaction in online customer service interactions. Two experiment scenarios were developed to test these links. Findings Overall, using a social-oriented communication style boosts customer satisfaction. Warmth perception of the chatbot mediates this effect, while chatbot’s anthropomorphised role (servant versus partner) moderates this effect. Originality/value This paper enriches the bilateral communication literature by extending the investigation on communication style effects to chatbot service interactions and revealing the psychological process driving the impacts. It also adds to the existing literature on chatbots as a customer service and contributes to the prominent topic examining how consumers react to artificial intelligence that is used to establish and maintain a relationship with them. Additionally, the authors also make contribution to the warmth and competence literature by demonstrating that customers can interpret social cues in chatbot service interactions mainly based on the warmth dimension. Thus, the authors further add to the growing chatbot humanness perception literature and respond to the calls for investigating more anthropomorphic design cues to enhance chatbot humanness. Finally, the authors also provide a way to help reconcile seemingly conflicting prior findings.