An Empirical Investigation into the Influence of Explanation Styles in Explainable AI on Consumer Acceptance: Evidence from E-commerce Platforms

With the widespread adoption of recommendation algorithms and growing concerns over algorithmic opacity, applying explainable AI (XAI) has become increasingly essential to mitigate algorithm-related anxieties. Drawing on the Heuristic-Systematic Model (HSM) and the Technology Acceptance Model (TAM), this study investigates how explanation styles in XAI (rule-based vs. example-based) interact with consumer decision domains (utilitarian vs. hedonic) to influence consumers’ adoption intentions. Perceived usefulness is introduced as a mediating variable, and need for cognition is examined as a moderating variable, a theoretical model of the matching effect between explanation style and decision domain is constructed in the context of an e-commerce platform. Three experiments were conducted to test the proposed model and hypotheses. The findings reveal that when consumers are in a hedonic decision domain, example-based explanations lead to higher adoption intentions than rule-based explanations. Conversely, in a utilitarian decision domain, rule-based explanations elicit higher adoption intentions than example-based explanations. Perceived usefulness mediates these effects. Furthermore, need for cognition moderates the effect of explanation–domain fit, rule-based explanations for utilitarian products enhance perceived usefulness among consumers with high need for cognition, while example-based explanations for hedonic products increase perceived usefulness and acceptance among consumers with low need for cognition. This research enriches the literature on XAI explanation styles in marketing and offers practical guidance for e-commerce platforms to effectively implement explainable AI strategies.