AI-DRIVEN PERSONALIZATION, AR VIRTUAL TRY-ON AND CONSUMER BEHAVIOR IN THE COSMETICS MARKET: THE MEDIATING ROLE OF CONSUMER TRUST

Abstract

Purpose. Digital transformation has moved artificial intelligence (AI) from the back office to the point of purchase. Nowhere is this clearer than in cosmetics, where algorithmic recommendations and augmented reality (AR) virtual try-on tools now mediate how consumers choose shades, formulations and brands. This study examines how two AI-enabled tools, AI-driven personalization (AIP) and AR virtual try-on (VTO), shape three consumer behavior outcomes: purchase intention (PI), customer satisfaction (CS) and brand loyalty (BL). It also tests whether consumer trust (CT) mediates these effects. Design/methodology. Drawing on the Stimulus–Organism–Response (S-O-R) framework, the study adopts a positivist, deductive, cross-sectional survey design. Data is collected from adult consumers in Pakistan's major cities (Lahore, Karachi and Islamabad/Rawalpindi) who have bought cosmetic products through an AI-enabled app, website or e-marketplace in the past six months. The proposed analysis in IBM SPSS Statistics covers reliability testing, exploratory factor analysis, correlation, multiple regression. Findings (illustrative). A statistical run on a simulated dataset (n = 384) that mirrors the proposed design shows that AIP and VTO both raise consumer trust (R² = .23) and all three outcomes. Trust significantly mediates all six indirect paths. Mediation is full for AIP → satisfaction and AIP → loyalty, and partial for the remaining paths. Originality. The study is among the first to model two distinct AI tools against three behavioral outcomes through a single trust mechanism in an emerging-market cosmetics context. It offers brands a trust-centered roadmap for AI investment.