Behavioral Study on Buyer Shortlisting in AI Mode

This report presents findings from a usability study examining how U.S.-based consumers shortlist major purchases using Google AI Mode compared to the standard Google Search experience. The study involved 48 participants who completed 185 tasks, utilizing a think-aloud protocol to capture their decision-making processes. Key findings indicate that 74% of shortlists in AI Mode were generated directly from AI outputs, with 64% of participants clicking nothing during their tasks. Trust dynamics differed significantly; AI Mode users relied heavily on AI-generated descriptions, while standard Search users sought corroboration from multiple sources. The study also found that unfamiliar brands were often overlooked, and participants rarely expressed concern about limited options in AI Mode. The methodology included remote usability testing and detailed analysis of participant interactions, providing insights into consumer behavior at the bottom of the purchase funnel. This research contributes to understanding the implications of AI on consumer decision-making.