The Impact of Google AI Overviews on Publisher Traffic and User Experience: Evidence from a Field Experiment

Abstract

Online search engines rely on external content to respond to user queries, while publishers depend on search engines for traffic. The integration of GenAI summaries into search interfaces, such as Google’s AI Overviews (AIOs), may reduce users’ incentives to visit downstream websites, yet causal evidence remains limited. We provide causal evidence by implementing a field experiment, using a custom Chrome extension, in which users are randomly assigned to either standard Google Search with AIOs or a version without them. Conditional on appearing, AIOs reduce outbound organic clicks by 39.8% and increase zeroclick searches by 34.5%, without affecting sponsored clicks or overall search frequency. Engagement on downstream websites, conditional on clicks, is not different across groups. Survey evidence shows no measurable improvements in perceived search quality or ease of finding information. Exploratory evidence from an AI Mode condition, in which users are redirected to a Gemini-powered conversational AI interface, suggests even greater reductions in traffic to downstream publishers. Overall, the results suggest that AIOs divert traffic away from publishers without improving the user experience or quality of engagement for website/

Keywords