AI-enabled delivery transparency and consumer repurchase intention in on-demand retail

Purpose This study examines how artificial intelligence (AI)-enabled delivery transparency (AEDT) influences consumer repurchase intention (RI) in on-demand retail by explaining how AI-generated fulfillment information is translated into perceived control (PC), consumer trust (CT) and repeat-purchase value. Design/methodology/approach This study develops a multi-theory framework based on the Technology Acceptance Model, Expectation-Confirmation Theory and Perceived Value Theory. It proposes a chain model of AEDT-perceived control-consumer trust-repurchase intention, with time sensitivity (TS) as a moderator. Data were collected from 507 on-demand retail consumers. Reliability and validity tests, confirmatory factor analysis, regression analysis, bootstrap mediation testing and moderation testing were used to examine the hypotheses. Findings AEDT significantly increases RI. PC and CT each mediate this relationship, and the chained pathway through PC and trust is also significant. Contrary to the initial hypotheses, TS negatively moderates the effects of delivery transparency, indicating a “high baseline-low marginal gain” mechanism among highly time-sensitive consumers. Research limitations/implications The cross-sectional design does not permit definitive causal inference or capture long-term behavioral dynamics across repeated purchases, service failures and recovery processes. The findings are primarily applicable to on-demand retail contexts in which consumers have experience with AI-enabled delivery information. Practical implications Platforms should treat delivery transparency as a service communication mechanism rather than merely a logistics tool. Highly time-sensitive consumers require stronger fulfillment assurance, exception handling, compensation and reliability guarantees, whereas less time-sensitive consumers may benefit more from visual tracking, progress explanations and dynamic feedback. Social implications This study suggests that AEDT can improve consumer experience by reducing uncertainty, strengthening trust and making delivery processes more visible and accountable. In on-demand retail, transparent AI logistics services may help platforms improve service fairness, reduce information asymmetry and enhance consumer confidence in digital retail ecosystems. The findings also imply that platforms should avoid treating consumers as homogeneous users and should provide differentiated information services for consumers with different levels of TS. Originality/value The study reframes AEDT as a consumer behavior and service marketing construct, identifies the chained psychological mechanism through PC and trust and reveals a reverse moderation effect that extends understanding of consumer heterogeneity in AI-enabled service transparency.