Definition
AI citation click-through rate is the share of times a user clicks a source link shown in an AI-generated answer, out of the times that answer was displayed. It is a narrower measure than the classic click-through rate, which counts clicks on a result link, because the click happens on a citation inside the answer rather than on a ranked listing.
The best-known public figure comes from Pew Research Center. Its analysis of 900 U.S. adults' browsing in March 2025 found that users who saw a Google AI summary clicked a link inside it in 1% of visits. Those users clicked a traditional result in 8% of visits, against 15% for users who saw no summary. This is specific to Google's AI summaries, that sample and that month, and says nothing directly about ChatGPT or Perplexity.
The low rate follows from how these answers work. They resolve the question on the page, which pushes behavior toward zero-click search. A citation can still shape a decision without a click: the user reads your brand name in the answer and later searches for you or types your URL. That effect shows up as AI dark traffic or branded search growth, so citation click-through understates influence, and a related problem is covered under zero-click attribution.
To calculate it for your own brand, divide referral visits from an AI surface by the number of times your page was cited there. The visits come from AI-referred traffic in your analytics or server logs, and the citation counts come from prompt tracking. Treat it as an estimate, since the number of times users actually saw a given answer is rarely known. Compare it across engines and page types instead of against the Pew figure.
For more context, see our review of AI search impact statistics, and our AI search visibility KPIs for where this metric fits next to mentions and citations.
Examples of AI Citation Click-Through Rate
- A SaaS company is cited in 3,000 tracked answers in a month and receives 45 referral visits from AI surfaces, an estimated 1.5% citation click-through rate, and uses it as a baseline for future months.
- A marketing lead cites Pew's 1% figure in a board deck and labels it as Google AI Overviews in March 2025 only, so nobody reads it as a ChatGPT benchmark.
- An analyst notices a how-to page is cited often but rarely clicked, while a pricing comparison page is cited less but clicked more, and moves effort toward comparison content.
- A brand sees flat AI referral visits but rising branded search after a citation push, and reports both numbers together instead of citation clicks alone.
