Enhancing the Effectiveness of Generative AI Travel Recommendations: Balancing Source Credibility and Cognitive Load

Abstract

Generative AI (GenAI) is revolutionizing the hospitality industry by offering recommendations via online travel agencies (OTA) to enhance customers’ decision-making. These suggestions’ perceived usefulness depends on both the content and how it is presented. We draw on cognitive load and source credibility theories in performing two studies to explore information structure and sources. We specifically examine how presentation formats (structured vs. unstructured), together with source disclosure (disclosed vs. undisclosed; Study 1) and source origin (internal vs. external; Study 2), influence GenAI recommendations’ perceived usefulness and consumers’ reuse intentions. Results indicate that structured formats reduce cognitive load, with effects varying by source disclosure. Undisclosed or external sources amplify the positive impact of structure, whereas internal sources substitute the structured format’s benefits with its inherent credibility. Our study extends cognitive load and source credibility theories to digital decision-making contexts while offering OTAs practical insights for designing more effective GenAI recommendations.