Generative engine optimization statistics measure how people use AI search, which sources AI engines cite, and what that does to clicks. This page collects the current figures, each one traced to the organisation that originally published it rather than to a secondary aggregator. Every statistic carries its source and collection date so you can check whether it still holds.
Last verified: 28 July 2026. This page is reviewed monthly. Figures older than twelve months are labelled as such.
As of July 2026, ChatGPT reports 900 million weekly active users and the Gemini app 950 million monthly active users, while 68.01% of US Google searches end without a click (SparkToro, Jan–Apr 2026). Earned media supplies 84% of the links AI engines cite, and paid content supplies 0.3% (Muck Rack, May 2026).
- 68.01% of US Google searches ended without a click between January and April 2026, up from 60.45% in 2024. (SparkToro / Similarweb clickstream panel, June 2026)
- Earned media accounts for 84% of all links cited by AI engines. Paid and advertorial content accounts for 0.3%. (Muck Rack, “What Is AI Reading?”, May 2026, 25M+ links)
- Brand web mentions correlate roughly three times more strongly with AI Overview visibility than backlinks do, across 75,000 brands studied. (Ahrefs, 2025)
- AI Overviews now appear on more than 20% of Google searches, and click-through rates fall by close to 60% when they do. (SparkToro, June 2026)
- Google AI Mode has passed 1 billion monthly active users. (Alphabet Q2 2026 earnings, 22 July 2026)
- ChatGPT reached 900 million weekly active users in February 2026, up from 400 million a year earlier. (OpenAI)
- Adding statistics to a page raised its visibility in generative engine responses by roughly 31% over an unoptimised baseline in controlled testing. (Aggarwal et al., KDD 2024)
- Distributing the same content through third-party publications produced a 239% median lift in AI search visibility, reaching 325% in the strongest cases. (Stacker / Scrunch controlled study)
How to read this page
Each statistic is written as a single extractable line: the figure, the claim, then the source and date in brackets. Where two credible studies disagree, both are shown with their methodology rather than averaged into one number. Figures sourced only to secondary blogs or AI-generated summaries have been removed from this page.
AI search adoption statistics
Adoption is the base rate underneath every other number here. If you want the definitions before the data, start with our guide to generative engine optimization.
How many people use AI search engines
ChatGPT weekly active users
OpenAI announced 900 million weekly active users in February 2026, more than double the 400 million reported in February 2025. Separately, Sensor Tower data reported by Reuters put the ChatGPT app past 1 billion monthly active users in June 2026.
Source: OpenAI, February 2026Gemini app monthly active users
Google reported 950 million monthly active users for the Gemini app, with daily active users tripling over the previous year. The app was at 750 million in Q4 2025 and roughly 400 million in mid-2025.
Source: Alphabet Q2 2026 earnings, 22 July 2026- 1 billion+ — monthly active users of Google AI Mode, which Google says is now sending billions of clicks to websites each week. (Alphabet Q2 2026 earnings, July 2026)
- 2 billion+ — monthly reach of Gemini-powered AI Overviews inside Google Search. (Google, 2026)
- 22 billion — API tokens per minute processed by Gemini models. (Alphabet Q2 2026 earnings, July 2026)
- 77% — share of US ChatGPT users who treat it as a search engine. (Adobe, July 2025 — over twelve months old, treat as directional)
- 80% — share of consumers relying on AI-generated summaries for at least 40% of their searches. (Bain & Company, February 2025 — over twelve months old, treat as directional)
How often AI Overviews appear in search results
Share of Google searches that return an AI Overview
AI Overviews appear on more than one in five US Google searches. When one is present, click-through rate to traditional results drops by close to 60%.
Source: SparkToro / Similarweb, January–April 2026AI Mode is a much smaller surface today but growing fast. SparkToro found that only 0.34% of searches transitioned into AI Mode during its January–April 2026 study window, while Google said at I/O 2026 that AI Mode query volume was more than doubling each quarter.
AI citation statistics
Citation data answers the question that matters most for generative engine optimization: when an AI engine builds an answer, whose pages does it pull from? Our own breakdown of source selection sits in the ChatGPT citations report.
Which sources AI engines cite most often
Muck Rack’s What Is AI Reading? study is the largest public dataset on this question. The May 2026 edition analysed more than 25 million links from ChatGPT, Claude, and Gemini responses across 17 industries.
| Source category | Share of AI citations | Stability across three study editions |
|---|---|---|
| Earned media (editorial, journalism, academic, government) | 84% | 82–89% since July 2025 |
| Journalism specifically | 27% | 25–27% since July 2025 |
| Paid and advertorial content | 0.3% | Negligible in every edition |
- 84% — share of AI citations coming from earned media. (Muck Rack, May 2026)
- 0.3% — share coming from paid or advertorial placements. (Muck Rack, May 2026)
- 27% — share coming from professional journalism, stable across all three editions of the study. (Muck Rack, May 2026)
- 3.5× — how much more often press releases appear in industry-trend answers than in “best of” list answers. (Muck Rack, May 2026)
- 25 million+ — links analysed in the May 2026 edition, across ChatGPT, Claude, and Gemini in 17 industries. (Muck Rack, May 2026)
The practical read: content you publish on your own domain is competing for a minority share of citations. The category that wins is coverage you earn somewhere else.
How brand mentions compare to backlinks as a citation signal
How much more strongly brand mentions correlate with AI Overview visibility than backlinks
Ahrefs found that unlinked brand web mentions were a substantially stronger correlate of AI Overview visibility than backlink counts. Correlation is not causation, but the gap between the two signals is large enough to change where link-building budget goes.
Source: Ahrefs, study of 75,000 brands- 239% — median lift in AI search visibility when the same content was distributed through third-party news outlets, across five leading LLMs. (Stacker / Scrunch controlled study)
- 325% — the upper end of that lift in the strongest cases. (Stacker / Scrunch)
Zero-click and click-through rate statistics
This is the number most often misquoted, so it is worth stating once, precisely, with its method attached.
Share of US Google searches that ended without a click, January–April 2026
SparkToro’s 2026 study measured US Google searches from January through April 2026, assuming two-thirds mobile and one-third desktop, and excluding searches made inside Google’s mobile search app, where zero-click behaviour may be higher still. The figure counts any search that produced no click at all, including clicks to Google-owned properties such as Maps and YouTube.
Source: SparkToro, using Similarweb desktop and mobile web panel data- 68.01% — US Google searches ending without a click, January–April 2026. (SparkToro, June 2026)
- +7.56 points — increase in the zero-click rate since 2024. (SparkToro, June 2026)
- −9.51 points — fall in the share of searches producing at least one click between 2024 and 2026, a 22.9% relative decline. (SparkToro, June 2026)
- +7.2 points — increase in searches that led to another Google search rather than an outbound click. (SparkToro, June 2026)
- 8% vs 15% — how often users clicked a traditional result with an AI Overview present versus without one. (Pew Research Center, 2025)
Why zero-click figures differ between studies
Different studies measure different things, which is why the published range runs from the high fifties to the high sixties. The comparison below is the honest version.
| Study and year | Zero-click rate | Data source |
|---|---|---|
| SparkToro, Jan–Apr 2026 (US) | 68.01% | Similarweb desktop and mobile panel |
| SparkToro, 2024 (US) | 58.5% | Datos clickstream panel |
| SparkToro, 2024 (EU) | 59.7% | Datos clickstream panel |
| SparkToro, 2020 | 64.82% | Similarweb |
| SparkToro, 2019 | 49–50.33% | Jumpshot (now defunct) |
SparkToro is explicit that these are not directly comparable across years because the underlying panels and providers changed. The direction is consistent; the absolute figures are not interchangeable. If a page quotes “60% zero-click” without naming a panel and a year, it is quoting a number that has moved.
Generative engine optimization effectiveness statistics
The most rigorous public test of which GEO tactics actually work is still GEO: Generative Engine Optimization by Aggarwal et al., presented at KDD 2024 by researchers from Princeton University, Georgia Tech, the Allen Institute for AI, and IIT Delhi. It tested nine optimisation methods across thousands of content samples and scored the results using position-adjusted word counts plus subjective impression scores.
| Method | Overall score | Subjective average |
|---|---|---|
| No optimization (baseline) | 19.3 | 19.3 |
| Keyword stuffing | 17.7 | 20.2 |
| Unique words | 20.5 | 20.4 |
| Easy-to-understand | 22.0 | 20.5 |
| Authoritative | 22.3 | 22.9 |
| Technical terms | 22.7 | 21.4 |
| Fluency optimization | 24.7 | 23.4 |
| Cite sources | 25.0 | 23.1 |
| Statistics addition | 25.4 | 23.7 |
| Quotation addition | 27.2 | 23.8 |
- 27.2 — highest-scoring method tested: adding relevant quotations, a 41% improvement on the unoptimised baseline of 19.3. (Aggarwal et al., KDD 2024)
- 25.4 — score for adding statistics, roughly 31% above baseline. (Aggarwal et al., KDD 2024)
- 17.7 — score for keyword stuffing, the only method that performed worse than doing nothing. (Aggarwal et al., KDD 2024)
- 19.3 — baseline score for content with no optimisation applied. (Aggarwal et al., KDD 2024)
Why this matters
Three of the four highest-scoring methods are about attribution: quoting sources, citing sources, and adding statistics. The tactic that mimics traditional SEO most closely, keyword stuffing, was the only one to score below baseline.
AI search traffic and market statistics
Traffic from AI engines is still small relative to organic search, but the composition of what remains is changing. If you are choosing a way to measure it, our comparison of AI visibility tools covers the current options.
- Billions of clicks per week — what Google states its AI features in Search send to websites. (Google, July 2026)
- 24% — Alphabet’s year-over-year revenue growth in Q2 2026, with Search and Other revenue up 17%. (Alphabet Q2 2026 earnings, July 2026)
- 27.4% — Gemini’s share of global AI chatbot web traffic as of mid-2026, up from single digits in early 2025. (Similarweb)
- 3× — growth in Gemini app daily active users over the twelve months to July 2026. (Alphabet Q2 2026 earnings)
How to use these generative engine optimization statistics
The pattern across every dataset on this page is consistent: AI engines reward attribution and third-party validation, and they resolve more queries without sending a click. If you are still building the internal case for the shift, the benefits of generative engine optimization covers what it delivers beyond citation counts. Five things follow from the data above.
- Track citations, not just rankings. If 68% of searches end without a click, click-based reporting is measuring a shrinking slice of your visibility.
- Shift budget toward earned coverage. Earned media supplies 84% of AI citations; paid placement supplies 0.3%.
- Add sourced statistics and direct quotations to pages you want cited. These were the two highest-scoring tactics in the KDD 2024 testing.
- Treat unlinked brand mentions as a ranking asset, not a vanity metric. They correlated three times more strongly with AI Overview visibility than backlinks.
- Re-verify any statistic older than twelve months before you publish it. Panels change, and so do the numbers.
Working through these in order is the fastest way to act on them. The generative engine optimization checklist expands the five items above into a full action plan.
Wellows tracks brand citations and visibility across ChatGPT, Gemini, Perplexity, Google AI Overviews, and Google AI Mode. See what the platform measures.
Methodology and sources
Every figure on this page was traced to the organisation that originally published it. Statistics previously sourced to secondary marketing blogs, or to pages that were themselves generated by AI assistants, have been removed rather than re-cited.
| Source | What it covers | Collection period |
|---|---|---|
| SparkToro | Zero-click rate, AI Overview frequency, AI Mode share | January–April 2026 |
| Muck Rack, “What Is AI Reading?” | Composition of AI citations across ChatGPT, Claude, Gemini | May 2026 edition, 25M+ links |
| Aggarwal et al., “GEO: Generative Engine Optimization”, KDD 2024 | Controlled testing of nine GEO methods | Published 2024 |
| Alphabet Q2 2026 earnings and Google I/O 2026 | Gemini, AI Mode, and AI Overviews usage | May and July 2026 |
| OpenAI announcements | ChatGPT weekly and monthly active users | February–June 2026 |
| Pew Research Center | Click behaviour with and without AI Overviews | 2025 |
| Ahrefs | Brand mentions versus backlinks as AI visibility signals | 2025, 75,000 brands |
| Bain & Company; Adobe | Consumer reliance on AI summaries | February and July 2025 |
Frequently asked questions about generative engine optimization statistics
