AI Personalization and Trust Formation in E-Commerce: Examining the Role of Experiential Mediators and the Privacy Paradox
TL;DR
This study tests Satisfaction and Engagement as parallel experiential mediators and Privacy Concerns as a boundary condition among 130 “digital-native” e-commerce users in India and offers managers a roadmap privileging experience design—especially engagement—over algorithmic sophistication.
Abstract
As artificial intelligence (AI) becomes central to e-commerce strategy, how algorithmic personalization translates into consumer trust remains contested. Drawing on the Stimulus-Organism-Response (S-O-R) framework, the Technology Acceptance Model (TAM), and Privacy Calculus Theory, this study tests Satisfaction and Engagement as parallel experiential mediators and Privacy Concerns as a boundary condition among 130 “digital-native” e-commerce users (aged 18–35) in India. Using hierarchical regression and percentile bootstrap mediation (5,000 resamples), AI personalization predicted trust in isolation (R² = .111, p < .001) but became non-significant once Satisfaction and Engagement entered the model (R² = .340, p < .001), supporting full mediation (total indirect effect B = 0.45, 95% CI [0.23, 0.63], p < .001). The Engagement pathway was robust (B = 0.20, 95% CI [0.10, 0.33], p < .001); the Satisfaction pathway was only marginal (B = 0.25, 95% CI [−0.01, 0.43], p = .064), making Engagement the more dependable mediator. Privacy Concerns showed no significant relationship with Trust (r = −.133, p = .131), evidencing a privacy paradox in which functional utility outweighs perceived data risk. The study extends TAM by identifying an “experiential bridge” for AI-driven trust formation and offers managers a roadmap privileging experience design—especially engagement—over algorithmic sophistication.