AI credibility and customer satisfaction: a neuroscience-based approach

Purpose The purpose of this study is to compare the impact of website recommendations and artificial intelligence (AI)-based, voice-enabled recommendations on customer satisfaction with the products recommended using the event-related potential (ERP) of electroencephalograms. Design/methodology/approach By applying source credibility theory and using experimental designs within subject factors, the study analysed the impact of AI and website recommendations based on the “search” and “experience” product types. Findings The behavioural results confirmed the significant effects of both AI and website recommendations on customer satisfaction with the products recommended. The ERP results showed that for a “search” product, users demonstrated a higher emotional response (represented by the P3 amplitude) to voice-enabled AI-recommended products and a higher attentional response (represented by the P1 amplitude) to products recommended by a website. However, the influence of the “experience” product type was ineffective. Originality/value The results contribute to the AI credibility literature by highlighting the effectiveness of AI-powered voice assistant recommendations on customer satisfaction.