Artificial intelligent chatbots as brand promoters: a two-stage structural equation modeling-artificial neural network approach.

Purpose: This study investigates the determinants of effective human and artificial intelligence (AI) relationship-building strategies for brands. It explores the antecedents and consequences of consumers' interactant satisfaction with communication and identifies ways to enhance consumer purchase intention via AI chatbot promotion. Design/methodology/approach: Microsoft Xiaoice served as the focal AI chatbot, and 331 valid samples were obtained. A two-stage structural equation modeling-artificial neural network approach was adopted to verify the proposed theoretical model. Findings: Regarding the IQ (intelligence quotient) and EQ (emotional quotient) of AI chatbots, the multi-dimensional social support model helps explain consumers' interactant satisfaction with communication, which facilitates affective attachment and purchase intention. The results also show that chatbots should emphasize emotional and esteem social support more than informational support. Practical implications: Brands should focus more on AI chatbots' emotional and empathetic responses than functional aspects when designing dialogue content for human–AI interactions. Well-designed AI chatbots can help marketers develop effective brand promotion strategies. Originality/value: This research enriches the human–AI interaction literature by adopting a multi-dimensional social support theoretical lens that can enhance the interactant satisfaction with communication, affective attachment and purchase intention of AI chatbot users. [ABSTRACT FROM AUTHOR]