Purpose This study aims to investigate how consumers' perceptions of artificial intelligence (AI) traits – warmth and competence – influence value-in-use across the customer decision journey (CDJ). Integrating the stereotype content model (SCM), cognitive fit theory (CFT) and the heuristic–systematic model (HSM), it examines the cognitive and emotional mechanisms shaping responses to AI at different decision stages. Design/methodology/approach Three experiments using a 2 × 3 factorial design (AI type × CDJ stage) were conducted in online grocery-shopping scenarios. Study 1 tested the direct effects of AI traits on value-in-use. Study 2 assessed perceived fit as a mediator. Study 3 examined heuristic versus systematic processing as a moderator. A robustness check used alternative stimuli. Findings Competence-based AI enhanced value-in-use during pre-purchase and purchase, whereas warmth-based AI was more effective post-purchase. Perceived fit mediated these effects, and processing mode moderated them: competence was more effective under systematic processing, while warmth performed better under heuristic processing. Originality/value This research extends SCM, CFT and HSM to AI-mediated retail by showing that the effects of AI traits depend on CDJ stages and processing conditions. It offers a framework for tailoring AI strategies to maximise value-in-use across the journey.
Lee et al. (Mon,) studied this question.