Gary, JostonXixing, ZhouTang, Yuk MingYang, GuMoreira, Antonio2026-01-172026-01-172025-11-281062-73751533-7995http://hdl.handle.net/10400.2/20965Retail practice shows that augmented-reality shopping applications with similar technical quality can elicit widely different consumer reactions. This study proposes a dual-pathway Stimulus–Organism–Response model: a technical pathway linking augmented realism, information richness, and personalization to interaction satisfaction, and an emotion-priming pathway where anticipated emotions shape immersion, telepresence, and pleasure without technical appraisal. Both converge at inspiration, the sole System-2 construct converting experience into behavior. Data from quasi-experimental participants were analyzed using PLS-SEM, SHAP-interpreted gradient boosting, and K-Means robustness checks. Information richness showed the strongest technical effect, while anticipated emotions most strongly affected immediate experiences. Inspiration predicted purchase and cross-buying intentions. Machine-learning diagnostics supported the framework and revealed non-linear thresholds in key pathways, clarifying inconsistent AR outcomes and positioning inspiration as the cognitive bridge to purchase.engAugmented realityAnticipated emotionsInspirationNonlinear explorationClustering algorithmjournal article2026-01-16cv-prod-462984010.4018/jgim.394246