This study examines the effects of anticipated negative evaluation, perceived ChatGPT credibility, and service cohesiveness on diners’ satisfaction through trust in ChatGPT’s competency. It further tests the moderating role of product fit diagnosticity, specifying the conditions under which generative AI strengthens customer experience in interactive marketing contexts.
A conceptual model grounded in the elaboration likelihood model (ELM) and human territoriality theory (HTT) was evaluated using survey data from 307 solo diners in Taiwan who had recently used ChatGPT for restaurant-related information. PLS-SEM was applied to analyse the structural relationships among the constructions, with particular attention to mediation and moderation mechanisms.
The results indicate that anticipated negative evaluation, credibility, and service cohesiveness significantly enhance trust in ChatGPT’s competency, which, in turn, predicts dining satisfaction with ChatGPT-assisted dining decisions. Product fit diagnosticity strengthens the trust-satisfaction relationship, while mediation analysis confirms trust as a central mechanism linking the antecedent variables to satisfaction.
This study contributes to interactive marketing research by integrating ELM and HTT into a unified framework, demonstrating how ChatGPT is perceived as a personalized companion associated with increased trust, reduced stigma and enhanced satisfaction.
