AI Search Visibility Statistics: Sourced Numbers for 2026

The most reliable AI search visibility statistics for 2026 show three things. Google users who see an AI summary click results about half as often (Pew, 8% versus 15%). AI referrals are under 1% of most sites' traffic. Where they convert, they convert far better, but not on every site.

Key takeaways

• Treat every AI figure as a result for one dataset, one date and one definition. None of the numbers below is a market average.
• Pew Research Center recorded 8% of visits clicking a traditional result when an AI summary appeared, against 15% when none did.
• Ahrefs reports AI search visitors converting at 23 times the rate of organic search on its own site. Microsoft Clarity reports 1.66% against 0.15% sign-up conversion across its publisher sample. A separate Clarity e-commerce analysis found the opposite for ChatGPT.
• AI referral traffic is small, so a conversion multiple on a tiny base can mislead. Compare absolute sign-ups, not only ratios.
• Citations move quickly. In Noma's own count of 2,545 answers, most cited URLs appeared in a single run, so one-off checks overstate stability.
• Rebuild the measures on your own brand with a fixed prompt set, a referral rule in analytics and the free AI SEO checker.

What are the latest AI search visibility statistics?

The table lists the figures worth quoting, each with its publisher, date and a limit on what it proves. Where a figure reached us through a news report of a study, the table says so. For the definitions behind these measures, see what AI visibility is.

Do ChatGPT visitors convert better than Google visitors?

Often yes on sign-ups, not reliably on purchases. Ahrefs found AI search visitors converted at 23 times the organic rate on its own site, while Microsoft Clarity's e-commerce analysis of 973 sites found ChatGPT referrals converting below every channel except paid social. The answer depends on your funnel.

Ahrefs' author describes the result in plain terms: "AI search visitors convert at a 23x higher rate than traditional organic search visitors for Ahrefs" (Ahrefs, June 2025). Note the last three words. It is one company, a software product with a free sign-up, measured over 30 days. The same post says bounce rate and pages per visit favoured AI visitors, but visit duration was worse, so the quality gain is mixed rather than clear.

Three studies, three definitions

The split is the finding. Sign-up and subscription goals reward people who arrive with a specific question already answered. Retail checkouts depend on price, stock and trust, which a chat answer does not settle.

Worked example: what the ratios mean in sign-ups

Take Clarity's two rates and 1,000 visits from each source. LLM referrals at 1.66% give about 16.6 sign-ups. Search at 0.15% gives 1.5. The ratio is 11 times, but the gap is 15 sign-ups per thousand visits. If AI sends 0.5% of a site's 100,000 monthly visits, as it did for Ahrefs, that is 500 visits and about 8 sign-ups at Clarity's rate. A high multiple on a small base is a signal to protect, not yet a budget case. The business case template in the AI visibility business case sets out the break-even arithmetic.

How much do AI Overviews reduce clicks?

Two independent sources point the same way, with different sizes. Pew measured real browsing; Ahrefs estimated click-through across keyword sets.

"Users who encountered an AI summary clicked on a traditional search result link in 8% of all visits." (Pew Research Center, July 2025)

Pew Research Center's comparison is 8% against 15%, a relative fall of about 47% ((15 − 8) ÷ 15). Pew also found that users ended their browsing session on 26% of pages with an AI summary against 16% without. Its method matched visits to the first page of results and the first three links cited in a summary, and it collected results from 7 to 17 April 2025.

Ahrefs' figure is larger. Per a May 2026 Ahrefs press release, AI Overviews were shown to have reduced clicks to top-ranking content by 58%. PPC Land's report adds the method: 150,000 keywords with an AI Overview against 150,000 informational keywords without, and an earlier April 2025 figure of 34.5%. The two datasets measure different things, so do not average them.

What this means for visibility: a page can rank first and lose most of its clicks. Being named or cited inside the summary is then the only exposure left for that query, which is why citation versus brand mention is worth separating in your reporting.

How much traffic do AI engines send?

Little, but growing fast. Microsoft Clarity reported AI-driven referrals up 155% over eight months, against 24% for search and 21% for social, and still under 1% of overall traffic (Microsoft Clarity, December 2025). The Ahrefs figure that ChatGPT sends about 190 times less traffic than Google, reported by PPC Land in February 2026, fits that picture.

Two readings follow. First, referral analytics undercount influence: someone who reads your name in an answer and later types your URL shows up as direct traffic. Second, a share of 0.5% to 1% means one month of data is noisy for small sites. Read quarterly trends. The AI search attribution guide covers the referral rules.

What do the numbers say about citations and engine behaviour?

Fewer public figures exist here, and they come with caveats. Three are worth knowing.

• Brand signals. Ahrefs' Q1 2026 AI Search Benchmark Report covers 13 studies, including 146 million search result pages and 730,000 AI responses. Its 75,000-brand study found YouTube mentions correlated more strongly with AI brand visibility than any other metric tested, with link volume and page count only weakly correlated (Ahrefs via Business Wire, May 2026). Correlation is not cause.
• Citation churn. In Noma by Nurdd's count of 2,545 answers from ChatGPT, Gemini, Perplexity and Claude, two in three cited URLs appeared in one run only, and 36% were cited again within a week. This is one prompt set and one period, not an industry benchmark.
• Engine disagreement. Noma's write-up of Bluefish's data reports that AI engines disagree on the top brand in 35% of buying topics. Tracking one engine tells you little about the others.

Noma by Nurdd's own measurement standard sets out numerators and denominators for these measures: see the AI Visibility Measurement Standard.

How do you read any AI search statistic? A five-question test

Apply this before quoting a number in a deck or an article. It takes two minutes per figure.

1. Whose data? A vendor's own site, a panel of users, or a prompt set the vendor wrote. Own-site data (Ahrefs' 23x) shows a possibility, not a norm.
2. What is the denominator? Visits, sessions, sign-ups, prompts or answers. A "58% drop" and a "47% drop" can both be right if the base differs.
3. What is "AI"? Chat assistants, AI Overviews, or both. Ahrefs defined AI search as assistants such as ChatGPT, Copilot and Gemini. Pew studied Google summaries only.
4. What date? Engines change products often. Pew's data is from March and April 2025; a 2026 reader should treat it as a baseline, not a current rate.
5. What outcome? Sign-up, purchase, citation or mention. Do not carry a conversion number across goal types, as the Clarity e-commerce result shows.

How do you check these measures on your own brand?

You can reproduce each number's logic with free tools and a spreadsheet before you pay for anything. Start with the page-level check, then move to prompts and referrals.

1. Check that engines can read you. Run the free AI SEO checker on your key pages. It shows whether ChatGPT, Claude, Perplexity and Google AI Mode can read the page and lists the top fixes. Then confirm bot access with the free AI crawler checker.
2. Build a fixed prompt set. Write 20 unbranded buyer questions and keep the wording unchanged. The free AI prompt finder helps with ideas. Guidance is in which prompts to track.
3. Run each prompt several times per engine. Record whether your brand is named, whether your domain is cited and which rivals appear. Mention rate is named answers divided by total answers. Repeat runs matter because of the churn above; see AI citation stability.
4. Add a referral rule in analytics. Segment sessions whose source is chatgpt.com, perplexity.ai, gemini.google.com or claude.ai. Report sessions, sign-ups and conversion rate per source, with the absolute counts beside the ratio.
5. Compare against search. Divide AI conversion by organic conversion on the same goal and period. If AI has under 50 sign-ups in the window, label the ratio as indicative.
6. Log the date and method. A statistic you cannot rerun is an anecdote.

Where Noma fits

Noma by Nurdd is an AEO platform that asks 13 AI engines, including ChatGPT, Gemini, Perplexity and Claude, a brand's buyer questions every day. For each engine it reports visibility, position and sentiment, the competitors named instead, and the sources each answer cites. It also reports which AI crawlers fetch each page and which assistants send visitors back.

See it on your own brand: book a demo or start the 3-day free trial. A card is taken at signup so billing can start on day 3; cancel before then and nothing is charged.

How we researched this

We reviewed primary pages (Pew Research Center's methodology and findings, Ahrefs' study post, Microsoft Clarity's blog and an Ahrefs press release) and trade reports of studies where the original was not available to us. Where we relied on a report, the table says so. We did not run our own conversion study, and the Noma figures come from Noma's published 2,545-answer analysis. Facts, figures and dates were checked on 2026-11-06. Several vendors' own roundups of AI statistics were also seen in search; we did not reuse their numbers unless we could trace them to a named source.

What to do next

Pick three figures from the table that matter to your decision, then reproduce them on your own data this week using the steps above. If you want that run daily across engines rather than by hand, book a demo or start the free trial. For the wider programme, read how to measure AI share of voice.