Unveiling Virtual Recommenders’ Experiences: The Differential Impact of Narrating Implicit and Explicit Experiences on Recommendation Effectiveness

Experiential narrative is a key strategy employed by marketers to facilitate product recommendations. With recent advances in generative artificial intelligence (GAI), virtual recommenders are now capable of narrating experiences. However, due to their non-human nature, the effectiveness of such strategy remains unclear, and existing findings are mixed. Drawing on the Strategic Experiential Modules, this research distinguishes between implicit (emotional and cognitive) and explicit (sensory, behavioral, and social) experiences, and integrates Language Expectancy Theory to develop a theoretical model that examines how virtual recommenders’ narration of different experiences influences consumers’ recommendation adoption intention. Across three experiments (N = 655), the findings reveal that the recommendation effectiveness of experiential narratives varies depending on the type of experience, and that language expectancy violation mediates this effect. Algorithm explanation plays a moderating role. By linking experience types, language expectancy violation, and recommendation adoption intention, this model provides a comprehensive and unified theoretical framework.