The impact of intelligent customer service agents’ initial response on consumers’ continuous interaction intention

Intelligent customer service agents (ICSAs) have been widely used as automated service assistants in e-commerce. Although ICSAs are highly valued by enterprises, consumers have lower continuous interaction intention for them. The initial response content will largely affect consumers' continuous interaction intention, which directly determines the ICSAs' efficiency. Therefore, making an effective strategy for ICSAs' initial response has become a focus for enterprises. Previous studies mainly investigated ICSAs from the perspective of the complete interaction, and ignored the initial interaction. Based on the theory of computers as social actors and the stereotype content model, this study investigates the impact of ICSAs' initial response content (functional content vs. social content) on consumers' continuous interaction intention. Through two online scenario-based experiments, the results show that initial response content has a significant influence on consumers' continuous intention. The initial response content impacts consumers' trust and affect through perceived warmth and perceived competence, which further influence consumers' continuous interaction intention. Furthermore, the functional response content has a greater impact on perceived competence, while the social response content has a stronger influence on perceived warmth. In addition, the effect of initial response content on perceived warmth and on perceived competence are moderated by interaction initiation modes. These findings enrich the literature on ICSAs and offer practical implications for e-retailers seeking to make effective initial response strategies.