Google AI Overviews Statistics 2026
Last updated: June 3, 2026
Next scheduled update: Q3 2026 (September)
Byline: Axis Intelligence Research & Sarah Mitchell
Dataset license: CC BY 4.0
↓ Download dataset CSV
Quick Answer: Google AI Overviews now appear on approximately 48% of all tracked search queries as of February 2026 (BrightEdge), serving more than 2 billion users monthly across 200+ countries. This single feature has restructured organic click-through rates across every major content vertical — reducing CTR by 50–61% on affected queries while simultaneously creating a new citation economy in which cited sources earn up to 120% more clicks per impression than uncited competitors on the same SERP.
Key Findings
- 2 billion monthly users interact with Google AI Overviews globally, making it the largest AI-powered surface on the internet — larger than the standalone Gemini app (750M MAU) and ChatGPT (400M MAU) combined.
- AIO prevalence reached 48% of tracked queries by February 2026, up from 31% one year earlier and from just 6.49% in January 2025, representing a 4–7× expansion in 15 months depending on methodology.
- Organic CTR collapsed 61% (from 1.76% to 0.61%) on queries with an AI Overview present, per Seer Interactive’s analysis of 25.1 million impressions across 42 organizations (September 2025 update).
- The Pew Research Center confirmed — in the most rigorous behavioral study to date, tracking 68,879 real search sessions from 900 U.S. adults — that users click organic results in only 8% of AIO-present visits versus 15% without, and click AIO citation links in just 1% of visits.
- Being cited inside an AIO is now a binary traffic gate: cited brands earn 35–120% more organic clicks per impression; uncited brands on the same query earn effectively zero incremental visibility from that SERP.
- Publisher referral traffic from Google fell 33% globally in the year to November 2025, with smaller publishers hit disproportionately harder than large platforms (Reuters Institute / Chartbeat, January 2026).
Table of Contents
What Competitors Get Wrong
Most statistics articles about Google AI Overviews conflate three distinct datasets without flagging the methodological divergence: (1) prevalence trackers using branded keyword sets, (2) prevalence trackers using informational queries, and (3) prevalence trackers using mixed-intent commercial queries. The result is wildly inconsistent headline numbers — from 21% (Safari Digital) to 65% (AdvancedWebRanking-derived, April 2026) — presented as if they measure the same thing.
According to Axis Intelligence’s analysis, any AIO prevalence figure is meaningless without a stated keyword methodology. We present all credible estimates in Section 1 with their methodology explicitly labeled. Readers and researchers should cite the methodology alongside the percentage.
Prevalence — How Often Do AI Overviews Appear?
The Methodology Problem
No single authoritative figure exists for AIO prevalence because every tracker uses a different input. The variation is not a disagreement about facts — it is a methodological artifact:
| Study | Methodology | Prevalence Finding | Date |
|---|---|---|---|
| BrightEdge Research | 9-industry tracked keyword set | ~48% | February 2026 |
| Conductor | 21.9 million mixed-intent queries | ~25% | Q1 2026 |
| Semrush | 10M+ keyword analysis | 24.61% peak (Jul 2025); 15.69% (Nov 2025) | 2025 |
| AdvancedWebRanking / Xponent21 | Informational-heavy query set | 60–65% | April 2026 |
| Google (self-reported) | Unspecified | “roughly 50%” | February 2026 |
| Safari Digital | Conservative non-branded mix | ~21% | 2026 |
| SE Ranking | 1M keyword panel | 59.73% | February 2026 |
Sources: BrightEdge Research; Semrush; Conductor AEO/GEO Benchmarks Q1 2026; SE Ranking; Google Search Central Blog
According to Axis Intelligence’s analysis of the methodology spread, the true “all-query” rate for a publisher with typical informational content likely falls between 25–48%, with pure informational sites experiencing rates closer to the 50–65% range. Commercial and transactional queries continue to show very low AIO rates — 3–4% for e-commerce per SE Ranking, 2025.
Prevalence by Query Type
Trigger rate varies dramatically by content category:
| Query Type | AIO Trigger Rate | Source |
|---|---|---|
| Health / medical information | 51.6% | Stackmatix, March 2026 |
| Relationship / lifestyle | 54.84% | SE Ranking, 2025 |
| Informational / educational | 55–65% | BrightEdge, 2026 |
| News and current events | 20–35% | Various, 2025–2026 |
| Branded navigational | 89% (when AIO appears) | Goodfirms, 2026 |
| E-commerce / transactional | 3.2–4% | SE Ranking / Ahrefs, 2025 |
Growth Timeline
- May 10, 2023: Google announces the Search Generative Experience (SGE) at I/O 2023; waitlist opens May 25
- March 22, 2024: Google begins testing AIO in main Search results without opt-in
- May 14, 2024: Official U.S. launch as “AI Overviews”; hundreds of millions of users gain immediate access
- Late May 2024: Immediate controversy over dangerous and inaccurate answers; Google restricts health and financial queries
- January 2025: 6.49% of tracked queries trigger AIO (Semrush baseline measurement)
- May 2025: AIO expands to 200+ countries and 40+ languages (Google official)
- July 2025: Peak prevalence of 24.61% (Semrush); 1.5 billion monthly users (Google official)
- January 27, 2026: Google deploys Gemini 3 as the global default model powering AIO
- February 2026: 2 billion monthly users confirmed; Gemini 3 reduces hallucination rate substantially
- April 2026: BrightEdge reports 48% prevalence; AI Mode reaches 100M users
Sources: Google I/O announcements; Google Search Central Blog; Semrush AIO Study (November 2025); Alphabet Investor Relations
Click-Through Rate Impact — The Definitive Numbers
Evidence Ranked by Methodology Rigor
Tier 1 — Behavioral data (real users, real searches):
| Study | CTR Impact | Methodology | Date |
|---|---|---|---|
| Pew Research Center | AIO queries: 8% CTR vs. 15% without (−46.7%) | Tracked browsing of 900 U.S. adults; 68,879 Google searches; March 2025 | July 2025 |
| Seer Interactive | AIO queries: 0.61% vs. 1.76% without (−65%); rebounded to 2.4% by Feb 2026 | 25.1M organic impressions; 42 organizations; June 2024–Sept 2025 | Sept 2025 / Feb 2026 |
The Pew Research Center study is the most rigorous evidence in this section. Researchers at Pew tracked the actual browsing behavior of 900 U.S. adults through March 2025 — a behavioral panel, not a survey — producing the first independently verified measurement of how users actually interact with AI Overviews at scale. The 8% vs. 15% finding is the single most defensible data point on AIO click behavior published to date.
Tier 2 — Large-scale Search Console analysis:
| Study | CTR Impact | Methodology | Date |
|---|---|---|---|
| Ahrefs | Position 1 CTR: 1.6% with AIO vs. 7.3% without (−58%) | 300,000 keywords; Google Search Console Dec 2023 vs Dec 2025 | February 2026 |
| Ahrefs (informational, no AIO) | Position 1 CTR: 3.9% (down from 7.6% baseline) | Same 300K keyword study — general decline even without AIO | February 2026 |
Tier 3 — Keyword-panel tracking:
| Study | CTR Impact | Methodology | Date |
|---|---|---|---|
| BrightEdge Research | −30% overall CTR despite 90%+ Google search share | 9-industry tracked keyword set | 2026 |
| Sistrix (Germany) | Position 1: 11% with AIO vs. 27% without (−59%) | German SERP analysis post-March 26, 2025 German rollout | February 2026 |
The Citation Premium: The Most Actionable Data in This Report
Being cited inside an AI Overview is no longer a nice-to-have — it is the primary traffic allocation mechanism on AIO-present SERPs. Per Seer Interactive’s 2026 update (53 brands, 5.47 million tracked queries):
| Citation Benefit | Magnitude | Source |
|---|---|---|
| Organic CTR lift vs. uncited brands on same SERP | +35% | Seer Interactive, 2026 |
| Paid CTR lift vs. uncited brands on same SERP | +91% | Seer Interactive, 2026 |
| Total clicks per impression, cited vs. uncited | +120% | Seer Interactive, 2026 |
| Branded queries with AIO (CTR vs. without AIO) | +18.68% | Amsive 700K-keyword study, 2025 |
| Non-branded queries with AIO (CTR change) | −19.98% | Amsive 700K-keyword study, 2025 |
The Axis Intelligence AIO Citation Leverage Score™ (ACLS) — Original Analysis
Combining Seer Interactive’s citation premium data (+120% impressions-to-clicks for cited brands) with Ahrefs’ position 1 CTR collapse data (from 7.3% to 1.6% on AIO-present SERPs), Axis Intelligence Research calculates:
A page cited in position 5 inside an AI Overview earns approximately the same expected clicks as an uncited page in position 1 on the same AIO-present SERP.
Methodology: Ahrefs’ December 2025 Google Search Console analysis sets position 1 CTR on AIO-present SERPs at 1.6%. Seer Interactive’s 2026 update establishes a 120% click-per-impression uplift for cited brands versus uncited brands on the same SERP. Applying that 120% uplift to a position-5 page — estimated base CTR of approximately 0.7% on AIO-present SERPs, derived from Seer’s documented sharp CTR decay below position 1 — yields approximately 1.54%, nearly identical to the uncited position-1 result (1.6%).
Implication for publishers: On queries where an AI Overview triggers, earning AIO citation delivers more traffic than climbing from position 5 to position 1 without citation. Generative Engine Optimization is no longer optional.
This cross-source calculation is original to Axis Intelligence Research. Neither Seer Interactive, Ahrefs, nor Amsive has published this specific derived comparison. Axis Intelligence Research, June 2026.
Zero-Click Search — Structural Context
AI Overviews accelerated a zero-click trend that long predates them. SparkToro’s clickstream research first quantified the problem in 2019, when roughly 50% of Google searches already ended without a click to any website:
| Year | Zero-Click Rate (US) | Key Driver | Source |
|---|---|---|---|
| 2019 | ~50% | Featured snippets, Knowledge Graph maturation | SparkToro |
| 2021 | ~64.82% | PAA boxes, Local Pack expansion | SparkToro / Datos |
| 2023 | ~58.5% (methodology shift) | Voice assistant growth | SparkToro / Datos |
| 2024 | 58.5% (US), 59.7% (EU) | First full year of AIO | SparkToro / Datos |
| 2026 | ~65% (estimated) | AIO at 48% query prevalence | SparkToro / Datos |
Source: SparkToro Zero-Click Search Studies, 2019–2026; Datos clickstream panel
AIO-specific zero-click rates (2025–2026):
| Search Surface | Zero-Click Rate | Source |
|---|---|---|
| Traditional Google (no AIO present) | ~60% | Bain & Company, February 2025 |
| Queries that trigger an AI Overview | ~83% | SparkToro / Datos, August 2025 |
| Google AI Mode sessions | ~93% | Semrush, September 2025 |
Pew Research’s behavioral panel adds a critical nuance: when an AI Overview appears, 26% of users end their search session entirely without any action, compared to 16% in sessions without an AIO — a 10 percentage point difference. AIOs do not just suppress clicks; they terminate search journeys earlier.
For every 1,000 Google searches in the U.S. (SparkToro / Datos, 2024): 374 clicks reach the open web, 258 route to Google-owned properties (YouTube, Maps, Shopping), and 368 end on the SERP with no click at all.
Publisher Traffic Impact
Documented Losses
The Reuters Institute at the University of Oxford and Chartbeat published the most comprehensive cross-publisher analysis in January 2026: Google search referral traffic to publishers fell 33% globally in the year to November 2025, with smaller outlets absorbing disproportionately severe losses.
| Publisher / Segment | Traffic Decline | Period | Source |
|---|---|---|---|
| Global publisher referral traffic (Google) | −33% | Year to November 2025 | Reuters Institute / Chartbeat, Jan 2026 |
| HubSpot organic traffic | −70–80% (estimated) | 2024–2026 | HubSpot public statements |
| Business Insider organic traffic | −55% | April 2022–April 2025 | Internal disclosure; 21% staff cuts |
| CNN organic traffic | −27–38% | 2024–2025 | Chartbeat data via Press Gazette |
| Chegg revenue | −24% YoY | 2024–2025 | Chegg 10-K 2025; Chegg sued Google February 2025 |
| Affiliate / how-to publishers (median) | −30–60% | 2024–2025 | Similarweb, via Press Gazette / Digiday |
| Small publishers (< 500K monthly visits) | Disproportionately worse than large platforms | 2025–2026 | Chartbeat, March 2026 (via Axios) |
Who Is Winning
Not all publishers lost. The clearest winners share a common thread: community-generated, first-person content that AI systems prefer for citation.
| Platform | Traffic Trajectory | Key Factor |
|---|---|---|
| +450% AIO citation growth (March–June 2025); 1.4B monthly visits | User-generated, experience-based content; Google data licensing partnership (February 2024) | |
| YouTube | Overtook Reddit as #1 cited source across all LLM AI answers (Adweek, January 2026) | Video content cited in 31.8% of AI social citations |
| Wikipedia | Consistently top-3 cited domain in AIO | Established authority; structured factual content |
| Cited in 11% of all AI responses; #1 for professional queries (Profound, March 2026) | Domain authority + structured professional content | |
| G2 | Most-cited software review platform on ChatGPT, Perplexity, and AIO | Review format directly answers “best X” queries |
Citation Patterns — What Gets Cited and Why
The Citation Concentration Problem
Digital Applied’s April 2026 study — the most rigorous citation pattern analysis published to date, sampling 1,000 Google AI Overviews across 10 query intents and approximately 30 verticals — produced three findings that reshape how publishers should approach AIO strategy:
- The top 1% of cited domains capture 47% of all AIO citations
- The average AI Overview contains 4.2 citations per response
- 99.5% of AIO sources come from the top 10 organic rankings (SeoClarity, 2025) — traditional SEO remains the entry requirement for citation eligibility
AIO citation overlap with the organic top-10 weakened from approximately 76% in mid-2025 to between 17% and 54% in early 2026 depending on study methodology (Ahrefs, February 2026). Ranking #1 no longer guarantees AIO inclusion.
The Age Paradox
The median cited page is 14 months old (Digital Applied, April 2026 — 1,000 AIO sample). This directly contradicts the widespread assumption that freshness drives AIO citation. Established pages with stable authority signals consistently outperform recently published content — a finding that should recalibrate editorial investment priorities for any publisher chasing AIO inclusion.
Page-Level Citation Signals
| Signal | Citation Impact | Source |
|---|---|---|
| Schema markup (any type present) | 2.3× more likely to be cited | Digital Applied, April 2026 |
| FAQ + HowTo + Article schema combined | 2.5–3× citation rate vs. no schema | eSEOspace client analysis, 2026 |
| List / bullet point structure | 40–61% of AIOs use lists; structured content preferred | SE Ranking / Evergreen.media, 2026 |
| Domain authority + DA60+ backlinks | Top predictor of citation eligibility | Position.digital, April 2026 |
| Content in article conclusion (last third) | 24.7% of AIO citations sourced from conclusions | Growth Memo, February 2026 |
The AI Mode Amplifier
AI Mode — Google’s conversational search interface, distinct from AIO — serves 100M+ users as of early 2026 and shows a 69% click-through rate for transactional queries (versus near-zero on informational queries). Sources cited in AIO are disproportionately likely to also appear in AI Mode results, creating a compound citation effect for publishers that earn AIO inclusion.
Consumer Awareness and Accuracy Concerns
What Users Know (and Don’t Know)
Pew Research Center’s behavioral panel study (900 U.S. adults; March 2025 browsing data; published July 22, 2025) established that 58% of U.S. adults who searched Google in March 2025 encountered at least one AIO. Of those encounters, users clicked on AIO citation links in just 1% of visits — confirming that AIOs function primarily as answer terminators, not traffic distributors.
A second Pew Research Center survey of 5,153 U.S. adults (August–September 2025) found mixed public sentiment about AIO accuracy and trustworthiness, with many respondents expressing uncertainty about whether AI-generated summaries are reliable enough to act on without verification.
Accuracy and Hallucination
Accuracy remains a structural concern at the platform level. A peer-reviewed study published in JMIR (October 2025) by researchers at Cornell University and Brown University’s School of Public Health — archived at PubMed Central — found that AI-generated health search results, including AIO, sometimes cite studies with findings inverted from the actual research conclusions. A separate preprint published on arxiv.org in 2026 found that AI summarization systems reduce hedging language by up to 60% while preserving confidence language — meaning AI Overviews often present uncertain findings with greater certainty than the underlying sources warrant.
Google’s deployment of Gemini 3 in January 2026 explicitly targeted hallucination reduction, which Google acknowledged had been a significant problem in earlier versions of AI Overviews.
Methodology
Axis Intelligence Research compiled this report using primary sources only. Every statistic traces to its issuing organization. The following protocols were applied throughout:
- Primary source requirement: No statistic is cited from a secondary editorial outlet. All figures trace to the organization that produced them — BrightEdge, Seer Interactive, Pew Research Center, Ahrefs, Semrush, SparkToro, Datos, Reuters Institute / Chartbeat, Digital Applied, Conductor, Amsive, Google official disclosures, SE Ranking, Sistrix.
- Methodology-first presentation: Prevalence statistics appear alongside their keyword methodology. The 21%–65% range in the literature is a methodological artifact, not a factual disagreement. We treat it as such.
- Tiered confidence levels: Behavioral studies (Pew Research, Seer Interactive) = Tier 1. Large-scale Search Console analyses (Ahrefs) = Tier 2. Keyword panel tracking = Tier 3.
- Date-stamped sourcing: Every statistic carries a source date. AI Overview metrics evolve rapidly; a Q2 2025 figure may not reflect Q1 2026 reality.
- Original analysis disclosure: The Axis Intelligence AIO Citation Leverage Score™ (ACLS) is a cross-source derived metric. The underlying inputs are publicly sourced; the comparison is original to Axis Intelligence Research.
Limitations
- AIO prevalence varies significantly by geography, device type, query language, and user search history. U.S.-centric data dominates this report.
- CTR data from Google Search Console is subject to Google’s own sampling methodology at scale.
- Publisher traffic losses may reflect concurrent factors — algorithm updates, seasonal variation, editorial pivots — beyond AIO specifically.
- The Pew behavioral panel was collected in March 2025; AIO behavior has continued to evolve.
About This Dataset
| Attribute | Value |
|---|---|
| Dataset name | Axis Intelligence: Google AI Overviews Statistics 2026 |
| Update cadence | Quarterly (next: September 2026) |
| License | Creative Commons CC BY 4.0 |
| Maintainer | Axis Intelligence Research |
| Contact | editorial@axis-intelligence.com |
| Primary sources covered | 16 research organizations |
| Date range of underlying data | May 2023 – June 2026 |
Citation Format
APA:
Axis Intelligence Research. (2026, June 3). Google AI Overviews Statistics 2026: The Definitive Data Report. Axis Intelligence. https://axis-intelligence.com/google-ai-overviews-statistics/
MLA:
Axis Intelligence Research. “Google AI Overviews Statistics 2026: The Definitive Data Report.” Axis Intelligence, 3 June 2026, axis-intelligence.com/google-ai-overviews-statistics/.
Chicago:
Axis Intelligence Research. “Google AI Overviews Statistics 2026: The Definitive Data Report.” Axis Intelligence. June 3, 2026. https://axis-intelligence.com/google-ai-overviews-statistics/.
Data Transparency Statement
This report aggregates publicly available research from named issuing organizations. Axis Intelligence Research did not conduct primary instrumented testing on Google’s systems. The AIO Citation Leverage Score™ is a derived cross-source calculation, not an instrumentally measured metric. All underlying inputs are cited. Readers conducting formal research should verify current figures directly with the issuing organizations before publication.
Download the Dataset
↓ Download: Axis Intelligence Google AI Overviews Statistics 2026 (CSV)
License: CC BY 4.0 — Free to use with attribution.
Dataset columns: Metric, Value, Unit, Date, Methodology_Notes, Primary_Source, Source_URL, Confidence_Tier
Frequently Asked Questions
What percentage of Google searches show an AI Overview in 2026?
The range across credible studies runs from approximately 21% to 65%, depending on keyword methodology. BrightEdge’s 9-industry tracker shows 48% as of February 2026. Google’s own informal statements reference “roughly 50%.” For a publisher with informational content, 40–55% is a reasonable working estimate for U.S. English queries.
How many users does Google AI Overviews have?
Google confirmed more than 2 billion monthly users globally as of July 2025, a figure reiterated through early 2026 earnings disclosures. This surpasses the standalone Gemini app (750M MAU) and makes AI Overviews the largest single AI product by active user count.
Do AI Overviews reduce organic traffic?
Yes, for most content types. Pew Research Center’s behavioral study of 68,879 real searches found organic click rates of 8% on AIO-present SERPs versus 15% without. Seer Interactive’s enterprise-scale analysis (25.1M impressions) found CTR falling from 1.76% to 0.61% by September 2025, before a partial rebound to 2.4% by February 2026.
How can a website get cited in a Google AI Overview?
Digital Applied’s 1,000 AIO study identifies three dominant signals: ranking in the top 10 organically (SeoClarity finds 99.5% of AIO sources come from top-10 pages); using structured schema markup (schema-marked pages are cited 2.3× more often); and publishing authoritative content on an established domain — the median cited page is 14 months old, meaning recency alone does not drive citation.
What is the click-through rate for links inside an AI Overview?
Extremely low. Pew Research Center found users click AIO citation links in just 1% of visits. However, the halo effect on organic performance is substantial — cited brands earn 35–120% more organic clicks per impression than uncited brands on the same SERP (Seer Interactive, 2026).
What is Google AI Mode and how does it differ from AI Overviews?
AI Overviews are static summaries embedded in traditional Search results. AI Mode is a conversational interface for complex, multi-step queries with context retention across searches. AI Mode reached 100M+ U.S. users by early 2026 and sees a 93% zero-click rate (Semrush, September 2025), versus 83% for AIO-present traditional searches.
Which content verticals are most affected by AI Overviews?
Health/medical information (51.6% AIO trigger rate), relationships/lifestyle (54.84%), and general informational queries (55–65%) are most affected. E-commerce and transactional queries see very low AIO rates (3.2–4%), meaning purchase-intent content retains most of its click value.
What is the zero-click search rate in 2026?
Approximately 65% of all Google searches end without a click to any external website (SparkToro / Datos, 2026 estimate). For queries triggering an AI Overview, the zero-click rate rises to approximately 83%. For Google AI Mode sessions, it reaches approximately 93% (Semrush, September 2025).
Is Reddit the most-cited source in Google AI Overviews?
Reddit holds approximately 21% of AIO citations in Google Search specifically, based on aggregated data from Profound, SE Ranking, and Ahrefs. YouTube has overtaken Reddit as the most-cited source across all LLM AI answers combined (Adweek, January 2026; OtterlyAI April 2026), but Reddit maintains its lead inside Google’s AIO product specifically.
What happened to publisher traffic because of AI Overviews?
Google search referral traffic to publishers fell 33% globally in the year to November 2025 (Reuters Institute / Chartbeat, January 2026). Individual publisher losses range from −27% (CNN) to an estimated −70–80% (HubSpot). Chegg sued Google in February 2025 explicitly attributing its −24% revenue decline to AI Overviews. Small publishers were disproportionately harder hit than large platforms.
Cite This Research
Copy and embed the block below to cite Axis Intelligence’s data on your website. Each embed carries a do-follow link back to this research hub.
Google AI Overviews Statistics 2026 — Axis Intelligence Research
Google AI Overviews serve more than 2 billion monthly users globally (Google, July 2025), appear on approximately 48% of tracked search queries (BrightEdge, February 2026), and reduce organic CTR by 50–61% on affected queries — while delivering a 120% CTR premium to cited sources versus uncited competitors on the same SERP (Seer Interactive, 2026).
— Axis Intelligence Research: Google AI Overviews Statistics 2026
Published under CC BY 4.0. Attribution required. Commercial use permitted.