AI SEO

AI Search Statistics: Overviews, Clicks and Source Visits

Explore verified AI search statistics from Pew Research Center, including AI Overview click behavior. Each figure includes its study population, original source, limits and downloadable data.

Last updated: Last verified:
Download the data (CSV) ↗7 source-checked findings

What users clicked after a Google results page, March 2025

No AI summary15%
With AI summary8%
AI source link1%
In Pew’s US browsing panel, 15% of visits without an AI summary led to a traditional result click; with a summary, 8% did. A cited source in the summary received a click in 1% of summary-page visits. These are observed associations, not a causal experiment.

The statistics at a glance

Swipe the table to see each value, year and original source.

AI Search Statistics: Overviews, Clicks and Source Visits: values, reporting years and original sources.
StatisticValueYearSource
#01Traditional result clicked when no AI summary appearedShare of Google results-page visits without an AI summary that led to a traditional result click, in a March 2025 panel of 900 US adults.15%2025Pew Research Center, Google AI summary browsing study ↗
#02Traditional result clicked when an AI summary appearedShare of Google results-page visits with an AI summary that led to a traditional result click, in a March 2025 panel of 900 US adults.8%2025Pew Research Center, Google AI summary browsing study ↗
#03AI-summary source link clicked when a summary appearedShare of visits to Google results pages with an AI summary where the next click went to a source cited in that summary; March 2025 US browsing panel.1%2025Pew Research Center, Google AI summary browsing study ↗
#04Observed Google searches that produced an AI summaryShare of 68,879 unique Google searches in Pew’s March 2025 US browsing dataset that produced an AI summary when results were collected in April 2025.18%2025Pew Research Center, Google AI summary browsing study ↗
#05Visits with an AI summary ending the browsing sessionShare of Google results-page visits with an AI summary after which the tracked browsing session ended; observational March 2025 US panel.26%2025Pew Research Center, Google AI summary browsing study ↗
#06Visits without an AI summary ending the browsing sessionShare of Google results-page visits without an AI summary after which the tracked browsing session ended; observational March 2025 US panel.16%2025Pew Research Center, Google AI summary browsing study ↗
#07US adults who said they had read AI search summariesShare of US adults reporting they had ever read an AI summary at the top of a search result; survey conducted February 17–23, 2026. Self-report, not observed browsing.60%2026Pew Research Center, Americans and AI 2026 survey ↗

The click gap needs a careful reading

In Pew Research Center's browsing analysis, US Google users clicked a traditional result after 15% of search-page visits without an AI summary, compared with 8% when one appeared. They clicked a source link inside the summary after only 1% of visits with a summary.

Those figures describe observed behavior, not a controlled experiment. Queries that trigger AI summaries may differ from queries that do not. The study does not prove that the summary alone caused the entire click gap, and it does not predict any individual publisher's traffic. It does show why being named as an AI source should not be treated as equivalent to receiving a visit.

How common were the summaries?

Pew found an AI summary for 18% of the 68,879 unique Google searches in its March 2025 browsing dataset, using search results collected the following month. Search results can change, and this sample was a panel of 900 US adults who agreed to share browsing data. The figure should not be presented as a current global rate for every search engine.

In a separate February 2026 Pew survey, 60% of US adults said they had ever read AI summaries at the top of search results. That is a self-reported survey response, not the same metric as observed search-page visits. Putting the two numbers side by side is useful only when their different methods are made clear.

What publishers can act on

Create pages that answer a specific question clearly, show where claims come from, and give the reader something a summary cannot: complete methodology, downloadable data, examples, or a usable tool. Measure actual search traffic and referral patterns over time. Do not claim that adding a FAQ block or structured data guarantees an AI citation.

The study also found that 26% of visits with an AI summary ended the tracked browsing session, compared with 16% without one. This is another association, not a forecast for every query. Query intent, result quality, and user needs all affect what happens next.

Source and limitations

Pew's study methodology explains the panel, result collection dates, and click classification. Cite the original research for its measurements; cite this guide when using our comparison chart or curated notes. For application, see our AI search content guide, clear answers guide, and AI SEO topic hub.

MAKE YOUR SOURCES EASY TO TRACE

How to cite this page

Choose a format and copy the reference. For numerical claims, always check the original linked source.

Online Marketing HQ. (2026, October 23). AI Search Statistics: Overviews, Clicks and Source Visits. https://onlinemarketinghq.co/stats/ai-seo-statistics

Chat on WhatsApp