Google AI Overview Statistics UK 2026

Statistics 21–30 — Click Behaviour, Zero-Click Search & Organic Traffic Impact

AI Overviews change more than the visual appearance of Google’s results page.

Independent clickstream and Search Console studies increasingly show that when a generated summary appears, users are less likely to click conventional organic listings.

The effect is not identical across every search.

But the evidence indicates that AI-generated answers are creating a measurable separation between:

• Search visibility.
• Organic rankings.
• Website clicks.

21. 18% of Google searches in Pew’s study generated an AI summary

The study examined real browsing behaviour rather than asking respondents to estimate how often they saw AI summaries.

This provides a useful behavioural benchmark, although prevalence will vary by market, query set and measurement date.

AI Overview implication: AI-generated summaries already affect a meaningful share of real-world search journeys.

Source: Pew Research Center, Google Users Are Less Likely to Click on Links When an AI Summary Appears in the Results, July 2025.

22. 58% of study participants encountered at least one Google AI summary

This reinforces the difference between the percentage of searches containing AI summaries and the percentage of users who encounter them.

A feature can appear on a minority of individual searches while still reaching a majority of active searchers over time.

AI Overview implication: User exposure can become widespread even before AI Overviews appear on most queries.

Source: Pew Research Center, 2025.

23. Traditional-result clicks fell from 15% to 8% when an AI summary appeared

The difference represents a substantial reduction in conventional result clicking within the study sample.

It does not mean every website loses the same percentage of traffic.

Impact can vary according to query type, ranking position, brand strength, SERP features and whether the generated answer satisfies the user completely.

AI Overview implication: Stable rankings do not necessarily guarantee stable organic traffic when the search interface changes above them.

Source: Pew Research Center, 2025.

24. Only 1% of visits to AI-summary pages produced a click on an AI-cited source

This finding is particularly important for publishers and organisations attempting to measure the value of being cited.

A citation can create visibility without producing proportional referral traffic.

AI Overview implication: AI citation visibility and website traffic need to be measured as separate performance dimensions.

Source: Pew Research Center, 2025.

25. 26% of visits to pages with an AI summary ended the browsing session

On search pages without an AI summary, the corresponding figure was 16%.

This suggests that generated answers can sometimes satisfy the information need sufficiently for the user to stop browsing.

AI Overview implication: Some searches that historically produced downstream website visits may increasingly conclude within Google’s own interface.

Source: Pew Research Center, 2025.

26. Around two-thirds of Google searches produced no click on a listed search result

This occurred whether or not an AI summary was present, showing that zero-click or non-result-click behaviour existed before AI Overviews became widespread.

AI-generated summaries should therefore be understood as accelerating a longer-term search-interface trend rather than creating zero-click search from nothing.

AI Overview implication: Organic traffic has never represented the total value of search visibility, and this distinction is becoming more important.

Source: Pew Research Center, 2025.

27. 88% of Google AI summaries cited at least three sources

Only around 1% cited a single source.

This shows that AI Overview visibility often operates as a multi-source environment rather than producing one conventional winner.

AI Overview implication: Organisations may compete for inclusion inside a generated answer even when they do not occupy the number-one conventional organic position.

Source: Pew Research Center, 2025.

28. The median Google AI summary was 67 words long

Individual summaries ranged from as few as 7 words to as many as 369 words.

Even a relatively short generated answer can occupy significant visual space above conventional search results.

AI Overview implication: The effect on organic visibility is not determined only by whether an AI Overview appears, but also by its size and the amount of screen space it occupies.

Source: Pew Research Center, 2025.

29. Ahrefs’ updated 2026 study found position-one CTR was 58% lower when an AI Overview appeared

The analysis compared:

• 150,000 informational keywords with AI Overviews.
• 150,000 informational keywords without AI Overviews.

It used aggregated Google Search Console data and compared performance with pre-AI Overview search behaviour.

Ahrefs had estimated a 34.5% reduction in its earlier 2025 study using the same broad methodology.

AI Overview implication: The traffic impact associated with AI Overviews may be increasing as the feature expands and user behaviour adapts.

Source: Ahrefs, Update: AI Overviews Reduce Clicks by 58%, February 2026.

30. CTR reductions were measured across all top-ten organic positions

The estimated impact by position was:

• Position 1: −58.0%.
• Position 2: −50.8%.
• Position 3: −46.4%.
• Position 4: −38.8%.
• Position 5: −32.6%.
• Position 6: −30.5%.
• Position 7: −29.7%.
• Position 8: −28.8%.
• Position 9: −29.7%.
• Position 10: −19.4%.

The largest effect was measured at the top of the organic results, where conventional CTR is normally strongest.

AI Overview implication: AI Overviews can change the economic value of ranking positions throughout page one rather than affecting only the first result.

Source: Ahrefs, 2026.

Statistics 31–40 — AI Overview Sources, Citations, Authority & Source Selection

AI Overview visibility creates a second competition inside Google Search.

A webpage can rank organically without being selected as an AI Overview source, while another page can be cited even when it does not rank prominently for the original query.

This distinction is becoming one of the most important changes in modern search optimisation.

31. Only 37.9% of AI Overview citations also appeared within Google’s first 10 result blocks

The analysis included conventional organic listings as well as SERP features such as:

• Featured snippets.
• People Also Ask.
• Video packs.
• Ads.
• Other search-result features.

This means the majority of URLs selected for an AI Overview were not drawn directly from the first ten visible result blocks for the original query.

AI Overview implication: Ranking prominently remains valuable, but page-one visibility alone does not guarantee selection as an AI Overview source.

Source: Ahrefs, Update: 38% of AI Overview Citations Pull From the Top 10, March 2026.

32. 31.2% of cited URLs appeared between result blocks 11 and 100

This shows that Google can draw sources from substantially deeper in the search ecosystem than the conventional first page.

AI Overview implication: Citation eligibility can extend beyond pages occupying the most visible conventional ranking positions.

Source: Ahrefs, March 2026.

33. 31.0% of AI Overview citations came from URLs outside the top 100 result blocks

This is one of the clearest pieces of evidence that AI Overview source selection cannot be explained purely by the conventional ranking order for the original search.

Google can expand the user’s search into related subqueries and retrieve supporting information from those additional result sets.

AI Overview implication: Organisations increasingly need topical coverage that can satisfy related subqueries, not only one primary keyword.

Source: Ahrefs, March 2026.

34. When only standard blue links were measured, 36.7% of cited URLs did not rank in the top 100

The corresponding distribution was:

• 37.1% ranked within the organic top ten.
• 26.2% ranked between positions 11 and 100.
• 36.7% did not rank within the top 100.

AI Overview implication: Conventional organic ranking and generative citation selection overlap, but they are clearly not identical systems.

Source: Ahrefs, March 2026.

35. 18.2% of cited pages that did not rank in the top 100 were YouTube URLs

This demonstrates the importance of formats beyond conventional webpages.

Google can extract information from video titles, descriptions, transcripts and related content when constructing generative answers.

AI Overview implication: AI-search optimisation increasingly extends beyond traditional text pages into video and other machine-readable content formats.

Source: Ahrefs, March 2026.

36. YouTube represented 5.6% of all AI Overview URLs cited in Ahrefs’ March dataset

This is significant because many of those YouTube pages were not prominent conventional organic results for the original search.

AI Overview implication: Video can contribute to generative visibility even when it is not the highest-ranking conventional result.

Source: Ahrefs, March 2026.

37. YouTube was the most-cited domain in Ahrefs’ September 2026 AI Overview dataset

Ahrefs measured more than 1.7 million distinct YouTube pages cited within the dataset.

The “mention share” metric represents a domain’s share of citations among the top 50 sources, not its share of every citation on the web.

AI Overview implication: Google’s generative search system is sourcing substantial amounts of information from video-based content.

Source: Ahrefs Brand Radar, The 50 Most-Cited Websites in Google AI Overviews, September 2026.

38. Reddit held an 18.5% mention share among the top 50 AI Overview sources

Other highly cited domains included:

• Facebook — 10.1%.
• Google — 8.8%.
• Instagram — 5.6%.
• Quora — 4.7%.
• Wikipedia — 4.0%.

The prominence of Reddit, YouTube, Facebook, Instagram and Quora shows that source selection is not limited to conventional editorial publishers.

AI Overview implication: Community discussion, video, user-generated content and social platforms can all contribute to the evidence Google uses when constructing answers.

Source: Ahrefs Brand Radar, September 2026.

39. Wikipedia, YouTube and Reddit collectively accounted for 15% of sources in Pew’s AI-summary sample

The same three domains represented approximately 17% of sources appearing in conventional Google results within the study.

This indicates that AI Overviews draw heavily from some of the same major information ecosystems as conventional Google Search, while assembling those sources differently.

AI Overview implication: Generative source selection represents an evolution of Google’s information ecosystem rather than an entirely separate web.

Source: Pew Research Center, July 2025.

40. Government websites represented 6% of AI Overview sources versus 2% of standard-result sources

Government sources were therefore represented at approximately three times the share within AI summaries in Pew’s sample.

News websites accounted for approximately 5% of sources in both AI summaries and conventional results.

AI Overview implication: Source type matters. For factual and public-information queries, authoritative primary sources may receive disproportionately strong generative visibility.

Source: Pew Research Center, 2025.

Statistics 41–50 — AI Mode, Generative Search Measurement, Business Adoption & the Future of Google Search

The final ten statistics move beyond AI Overviews themselves and examine the wider search ecosystem developing around them.

Google is increasingly combining generative answers, conversational queries, images, voice and multimodal search within the same search platform.

At the same time, generative AI adoption among UK consumers and businesses is increasing rapidly.

41. AI Mode queries are approximately three times longer than traditional searches

At AI Mode’s UK launch in July 2025, Google had described early-user questions as two to three times longer than conventional searches.

The newer data points towards increasingly detailed and conversational search behaviour.

AI search implication: Keyword-level optimisation alone becomes less descriptive when users express complete problems, constraints and preferences within a single query.

Source: Alphabet / Google, Q4 2025 remarks; Google UK AI Mode launch.

42. Nearly one in six AI Mode queries are now non-text

This illustrates how generative search is becoming multimodal.

Users can increasingly search by:

• Text.
• Voice.
• Images.
• Camera input.
• Combinations of these formats.

AI search implication: Images, video, structured product data and other visual assets may become increasingly important components of search visibility.

Source: Alphabet / Google, Q4 2025 remarks.

43. Circle to Search is available on more than 580 million Android devices

Circle to Search allows users to search directly from visual content displayed on their phone rather than beginning with a conventional typed query.

AI search implication: Search discovery increasingly begins from objects, images and visual context as well as words typed into a search box.

Source: Alphabet / Google, Q4 2025 remarks.

44. 59% of UK adults used generative AI within the previous three months

This comes from DSIT’s nationally representative Public Engagement Survey 2025/2026.

Usage reached at least 75% among adults below the age of 45.

AI search implication: Generated answers are becoming familiar to a majority of UK adults rather than remaining an early-adopter behaviour.

Source: Department for Science, Innovation and Technology, Public Engagement Survey 2025/2026.

45. 56% of UK adults recently used AI capable of generating human-like text or speech

This category directly overlaps with many conversational search and generative-answer experiences.

AI search implication: The interaction model used by AI search is already familiar to more than half of UK adults.

Source: DSIT, Public Engagement Survey 2025/2026.

46. 54% of UK adults recently used AI-powered digital assistants that understand natural language

This provides additional evidence that conversational interfaces are becoming part of everyday digital behaviour.

AI search implication: Search behaviour is moving progressively away from short keyword strings towards natural-language interaction.

Source: DSIT, 2026.

47. 74% of UK adults are concerned that AI-generated information may be inaccurate

At the same time, 47% identified easier access to information or advice as a benefit of AI.

This creates an important tension between convenience and trust.

AI search implication: Authoritative, transparent and verifiable sources may become increasingly valuable as users seek confidence in AI-generated information.

Source: DSIT, Public Engagement Survey 2025/2026.

48. AI use among UK businesses with 10 or more employees has risen to around 35%

The corresponding figure was approximately 12% in late 2023.

Business adoption has therefore expanded substantially in less than three years.

AI search implication: The organisations competing for visibility are themselves becoming increasingly AI-enabled, increasing expectations around automation, research and generative discovery.

Source: Office for National Statistics, Artificial Intelligence in UK Businesses: 2023 to 2026.

49. AI adoption reaches 49% among UK businesses with more than 250 employees

Adoption is therefore substantially higher among larger organisations.

AI search implication: Enterprise search strategy is increasingly developing inside organisations that are simultaneously adopting AI in other operational and commercial functions.

Source: Office for National Statistics, September 2026 thematic AI account using BICS data.

50. 17% of UK businesses with 10+ employees were using large-language-model text generation by June 2026

The proportion had increased by approximately 12 percentage points since the question was first introduced in September 2023.

Visual-content creation was the next most frequently reported technology at 14%.

AI search implication: Large language models are becoming embedded in business workflows at the same time as large language models increasingly mediate customer discovery through search.

Source: Office for National Statistics, Measuring Artificial Intelligence in the UK Economy Using a Thematic Account, September 2026.

CGO Media Analysis — What the 50 Google AI Overview Statistics Tell Us

Taken together, the 50 statistics show that Google AI Overviews are not simply another SERP feature.

They represent a structural change in how information is retrieved, synthesised, displayed and attributed within Google Search.

The result is a search environment in which conventional ranking visibility and generative citation visibility increasingly operate alongside one another.

1. AI Overviews Have Become a Mainstream Search Layer

The first conclusion is scale.

Google already operates inside an enormous UK search market, and AI Overviews are appearing across a meaningful proportion of those searches.

The feature no longer belongs to a limited test environment.

For businesses, publishers and organisations, this means generative visibility needs to be considered within normal search strategy rather than treated as a separate experimental project.

2. Informational Search Is the Most Exposed

The triggering data shows a particularly strong relationship between AI Overviews and informational intent.

Queries asking Google to:

• Explain.
• Define.
• Compare.
• Reason.
• Answer a question.
• Provide instructions.

are substantially more likely to generate an AI Overview than many short navigational or local searches.

This matters because informational content has historically been a major source of organic discovery and upper-funnel traffic.

3. Longer Queries Create Greater AI Overview Exposure

The relationship between query length and AI Overview appearance is one of the clearest findings in the dataset.

As searches become longer, they generally contain more context and more clearly expressed requirements.

That creates a natural environment for generative systems to synthesise information across multiple sources.

4. Commercial Search Is Beginning to Change as Well

AI Overviews were initially associated most strongly with informational searches.

The newer commercial-intent evidence suggests that this boundary is moving.

Generated answers are increasingly appearing during research and comparison stages of the buying journey.

This can affect queries such as:

• Best provider.
• Best software.
• Product comparisons.
• Service comparisons.
• Recommendations.
• Alternatives.

These queries are commercially important because they often occur closer to supplier or product selection than purely informational research.

5. Local Search Remains Structurally Different

Local-intent searches currently show substantially lower AI Overview triggering rates than many informational queries.

That makes sense because local searches often require immediate geographic results rather than broad synthesis.

For a query such as “dentist near me”, the user may need:

• Distance.
• Opening hours.
• Reviews.
• Directions.
• Availability.

rather than a long generated explanation.

Google Maps, local packs and Google Business Profile therefore remain particularly important for local visibility.

6. Organic Ranking and Organic Traffic Are Separating

One of the most significant findings is that a stable ranking does not guarantee a stable click-through rate.

Independent behavioural and Search Console studies indicate that AI Overviews can reduce the number of users clicking conventional results.

That changes the interpretation of ranking reports.

A page may remain in position one while receiving materially fewer clicks than it did before AI-generated answers became common for that query.

7. Citation Visibility Does Not Automatically Produce Referral Traffic

Pew’s clickstream evidence shows that users rarely click the links cited inside an AI-generated summary.

This creates a new measurement problem.

A business can gain:

• Brand exposure.
• Source attribution.
• Authority visibility.
• Recommendation presence.

without receiving an immediate website visit.

Citation visibility and referral traffic therefore need to be treated as different metrics.

8. Zero-Click Search Is Becoming More Important

Zero-click behaviour existed before AI Overviews.

Search engines have long displayed:

• Featured snippets.
• Knowledge panels.
• Weather answers.
• Calculators.
• Maps.

AI Overviews extend that model by answering more complex information needs directly inside Google.

This means website traffic increasingly represents only one part of the value created through search visibility.

9. The Conventional Top Ten Does Not Control AI Overview Citations

The citation research shows significant overlap between strong rankings and AI Overview source selection.

But the overlap is far from complete.

A large share of cited URLs come from:

• Positions below the conventional top ten.
• Pages outside the top 100.
• Different SERP features.
• Related queries produced by query fan-out.

This means AI Overview citation eligibility cannot be reduced to “rank first and you will be cited”.

10. Query Fan-Out Changes the Optimisation Target

Google’s use of query fan-out is central to understanding modern generative search.

A user may submit one query, but Google can explore multiple related questions and subtopics before constructing the answer.

The optimisation target therefore becomes broader than one keyword.

This rewards broader knowledge architecture and topic coverage.

11. Citation Authority Is Becoming a Separate Visibility Layer

Traditional SEO largely asks:

Can this page rank?

Generative search adds another question:

Can this source be selected and cited?

These are related questions, but the evidence suggests they are not identical.

This is why citation authority is becoming a useful concept alongside ranking authority.

12. Source Diversity Is Greater Than Conventional SEO Models Suggest

The source data shows substantial visibility for:

• YouTube.
• Reddit.
• Facebook.
• Instagram.
• Quora.
• Wikipedia.
• Government websites.

This demonstrates that Google’s generative layer can draw from a much wider evidence ecosystem than conventional editorial webpages alone.

Content strategy therefore increasingly needs to consider where credible information about the organisation exists across the wider web.

13. Video Has Become a Significant Generative Source Format

YouTube’s prominence in citation studies is especially notable.

Video should no longer be treated only as a separate social or brand channel.

Transcripts, descriptions, titles and contextual information can make video content available to generative retrieval systems.

14. Primary and Institutional Sources Can Carry Disproportionate Value

The higher representation of government sources within Pew’s AI-summary data suggests that factual and authoritative source types can play a strong role in generated answers.

For organisations, this reinforces the value of:

• Original research.
• Primary data.
• Clear methodology.
• Transparent authorship.
• Strong institutional identity.

Content that merely repeats existing commentary may have less distinctive source value than content introducing original evidence.

15. AI Search Is Becoming More Conversational

Google reports that AI Mode queries are substantially longer than traditional searches.

Users can express complete requirements rather than reduce them to short keyword phrases.

For example:

“best CRM software”

can become:

“Which CRM is best for a UK professional-services business with 50 employees, recurring contracts and integration with Microsoft 365?”

This increases the importance of detailed and explicit content.

16. Search Is Becoming Multimodal

Voice, images and visual search are becoming increasingly integrated into Google’s discovery environment.

This means search visibility increasingly involves:

• Text.
• Images.
• Video.
• Product data.
• Structured information.

Search optimisation can no longer be viewed entirely through text pages and keyword rankings.

17. AI Adoption Is Changing Customer Expectations

Generative AI is already familiar to a majority of UK adults.

As users become accustomed to generated answers, they are likely to expect:

• More direct answers.
• More detailed responses.
• Better comparisons.
• More conversational interaction.

Traditional search experiences are therefore evolving alongside wider AI behaviour rather than in isolation.

18. Trust Becomes More Important as AI Usage Increases

The UK evidence reveals an important contradiction.

AI adoption is increasing rapidly, but concern about inaccurate information remains high.

That creates an opportunity for organisations capable of demonstrating:

• Expertise.
• Evidence.
• Methodology.
• Accountability.
• Clear sourcing.

Trust signals become particularly important when generative systems combine information from multiple external sources.

19. AI Overview Visibility Is Becoming More Measurable

The introduction of dedicated generative-search reporting inside Search Console is strategically significant.

For the first time, organisations can increasingly distinguish generative visibility from conventional search performance using first-party Google data.

This should move AI search measurement away from speculative visibility claims and towards measurable performance analysis.

20. SEO and GEO Are Converging

The 50 statistics do not support the idea that SEO and generative optimisation should become completely separate disciplines.

Google itself continues to emphasise the importance of conventional SEO foundations.

The overlapping requirements include:

• Crawlable content.
• Strong information architecture.
• Useful content.
• Entity clarity.
• Authority.
• Evidence.
• Accurate structured information.

Generative optimisation adds citation and source-selection considerations on top of those foundations rather than replacing them.

The Strategic Shift

The conventional SEO objective was approximately:

Keyword → Ranking → Click

The emerging model is broader:

Topic → Entity → Evidence → Ranking + Citation → Click / Influence

What UK Businesses Should Do in 2026 to Improve Google AI Overview Visibility

There is no single technical switch that makes a website appear inside Google AI Overviews.

Google’s own guidance continues to emphasise the same core foundations required for conventional Search: content must be accessible, indexable, useful and eligible to appear in Google Search.

The difference is that generative search can evaluate information across multiple related queries and combine evidence from several different sources.

That changes the strategic objective from optimising one page for one keyword to building sufficient authority and information coverage around the underlying topic.

1. Maintain Strong Conventional SEO Foundations

Google AI Overviews remain part of Google Search.

That means websites should first ensure that Google can efficiently:

• Crawl important pages.
• Render the content.
• Understand internal linking.
• Identify canonical URLs.
• Index the correct pages.
• Access important text, images and video.

An organisation cannot reasonably expect consistent generative visibility if its core search infrastructure is weak.

2. Build Topic Coverage, Not Isolated Keyword Pages

Query fan-out means Google may explore related questions before constructing an AI Overview.

A business should therefore build connected content covering:

• The main topic.
• Definitions.
• Use cases.
• Problems.
• Comparisons.
• Alternatives.
• Implementation.
• Costs.
• Limitations.
• Industry-specific applications.

This creates a broader evidence environment than a single page targeting one phrase.

3. Answer the Questions Around the Query

The triggering statistics show that question-based searches are particularly likely to produce AI Overviews.

Businesses should identify the natural questions customers ask before, during and after a purchase decision.

Useful content should answer questions such as:

• What is it?
• How does it work?
• Who is it suitable for?
• What does it cost?
• What are the alternatives?
• What are the advantages and limitations?
• How does it compare?

Answers should be direct enough to extract but detailed enough to demonstrate genuine expertise.

4. Publish Original Evidence

Content that merely repeats information already available across hundreds of websites provides limited source differentiation.

Businesses should create evidence that originates with the organisation where possible.

Examples include:

• Original research.
• Surveys.
• Datasets.
• Industry statistics.
• Benchmarks.
• Case studies.
• Testing.
• Technical analysis.

Original evidence gives other publishers, journalists, search engines and generative systems a reason to reference the organisation directly.

5. Make Evidence Easy to Identify

Important claims should be presented with enough context for both users and search systems to understand where the information comes from.

Research-led pages should clearly identify:

• The statistic or finding.
• Source.
• Date.
• Geography.
• Sample or dataset where relevant.
• Methodology.
• Limitations.

This also improves the usefulness of the content to journalists and researchers.

6. Strengthen Organisational and Author Identity

Google needs to understand who is responsible for the information it encounters.

Businesses should make organisational identity explicit through:

• Detailed About pages.
• Research-team pages.
• Author pages.
• Editorial or research policies.
• Contact information.
• Consistent organisation naming.

For high-trust topics, transparent authorship and organisational accountability are particularly important.

7. Develop Strong Internal Knowledge Architecture

Related research, services, frameworks, statistics and supporting content should not exist as isolated pages.

Internal links should make relationships clear.

A useful architecture may connect:

This helps users and search systems understand how individual pages contribute to the wider subject.

8. Use Structured Data Where It Accurately Describes the Page

There is no special “AI Overview schema”.

However, supported structured data can still help Google understand what a page represents.

Depending on the content, relevant markup may include:

• Article.
• Organization.
• Dataset.
• Product.
• LocalBusiness.
• VideoObject.
• BreadcrumbList.

Structured data should accurately represent visible page content and follow Google’s guidelines.

It should not be treated as a shortcut to AI Overview inclusion.

9. Invest in Video

YouTube’s prominence in AI Overview citation studies makes video increasingly important to generative search strategy.

Businesses with genuine expertise should consider producing videos that explain:

• Processes.
• Products.
• Research findings.
• Industry questions.
• Comparisons.
• Technical subjects.

Titles, descriptions and transcripts should accurately describe the information contained within the video.

10. Improve Image and Multimodal Search Readiness

Google is increasingly providing first-party reporting for image-led and multimodal discovery.

Businesses should therefore improve:

• Image quality.
• Descriptive filenames.
• Alt text.
• Relevant surrounding text.
• Image metadata where appropriate.
• Product imagery.
• Diagrams and original figures.

Images should contribute useful information rather than exist only as decoration.

11. Build Independent External Authority

The source-selection evidence shows that Google’s generative system considers information from across the wider web.

Businesses should therefore seek legitimate external visibility through:

• Editorial coverage.
• Industry publications.
• Academic references.
• Professional associations.
• Trade journals.
• Research repositories.
• Authoritative directories.

The objective is not artificial link accumulation.

It is to establish independent evidence that the organisation and its expertise are recognised outside its own website.

12. Treat Digital PR as Part of AI Search Strategy

Digital PR can support both conventional SEO and generative visibility when it generates independent coverage of genuine research, expertise or business activity.

Strong digital PR assets can include:

• Original statistics.
• Research findings.
• Industry commentary.
• Data-led reports.
• Expert analysis.

This can distribute consistent information about the organisation across sources that search and AI systems may encounter independently.

13. Keep Important Information Consistent Across the Web

Conflicting information creates uncertainty.

Core organisational facts should remain consistent across:

• The company website.
• Profiles.
• Directories.
• Research repositories.
• Press coverage.
• Social platforms.

This is especially important for names, locations, services, leadership and other factual organisational information.

14. Use Google’s Preferred Sources Feature Where Relevant

Google’s Preferred Sources feature now extends into AI Overviews and AI Mode in supported markets.

For publishers and research-led websites, audiences can select a site as a preferred source.

Where appropriate, organisations can make it easier for regular readers to select the publication or research domain as one of their preferred sources.

This should be treated as an audience-loyalty feature rather than a substitute for organic authority.

15. Monitor Generative Search in Search Console

Generative search should no longer be measured entirely through third-party monitoring tools.

Google now provides dedicated reporting that can help organisations analyse:

• Generative-search impressions.
• Landing pages.
• Countries.
• Devices.
• Performance trends.

Businesses should benchmark visibility over time and compare it with conventional organic performance.

16. Monitor Multimodal Search Separately

Search Console’s new multimodal reporting can surface discovery generated through:

• Google Lens.
• Circle to Search.
• Image uploads.
• Chrome image search.

This is particularly relevant for:

• Ecommerce.
• Travel.
• Property.
• Fashion.
• Food.
• Automotive.
• Visual products.

17. Do Not Measure Success Through Citations Alone

Being cited is useful evidence of generative visibility.

But a citation is not automatically a commercial outcome.

Businesses should also monitor:

• AI referral traffic.
• Organic traffic.
• Branded search.
• Direct traffic.
• Leads.
• Conversions.
• Revenue.

This helps distinguish visibility from actual commercial impact.

18. Protect High-Value Commercial Pages

As AI Overviews expand further into commercial research, businesses should monitor queries surrounding:

• Best providers.
• Comparisons.
• Alternatives.
• Reviews.
• Costs.
• Recommendations.

These searches often influence the shortlist before a customer reaches a direct transactional query.

19. Avoid Producing Large Volumes of Generic AI Content

Generative search increases the temptation to scale content production aggressively.

Volume without additional information value can create a large website without creating a strong source.

Content should add something useful through:

• Experience.
• Original analysis.
• Data.
• Research.
• Examples.
• Expertise.

20. Build for Both Rankings and Citations

The strongest strategy is not to choose between SEO and generative optimisation.

Build content capable of doing both.

The Practical Priority

Businesses should resist chasing speculative “AI Overview hacks”.

The strongest evidence currently points towards strengthening the same qualities that make an organisation a useful source in any information ecosystem:

• Accessibility.
• Clarity.
• Depth.
• Original evidence.
• Authority.
• Trust.
• External recognition.

Google AI Overview Measurement Framework

AI Overview performance should not be measured with a single metric.

A useful measurement system needs to distinguish between conventional ranking visibility, generative visibility, citations, referral traffic, brand influence and eventual commercial outcomes.

1. Track Generative Search Impressions

Use Google’s generative-search reporting to establish whether pages are appearing inside AI-powered Search experiences.

Monitor:

• Total generative-search impressions.
• Month-on-month change.
• Year-on-year change where available.
• Pages receiving generative impressions.
• Country and device distribution.

This provides a first-party baseline for how visible the site is within Google’s generative layer.

2. Measure Citation Frequency

Citation monitoring should answer:

• How often is the brand cited?
• Which pages are cited?
• Which topics trigger citations?
• Which competitors are cited alongside the organisation?

A rising citation rate can indicate growing source eligibility even before referral traffic increases materially.

3. Separate Ranking Visibility From Citation Visibility

Because ranking and citation systems overlap imperfectly, report them separately.

For priority queries, record:

• Organic ranking position.
• Whether an AI Overview appears.
• Whether the organisation is cited.
• Whether the cited page matches the ranking page.

This helps identify pages that rank but are not cited, and pages that are cited despite weaker conventional rankings.

4. Track AI Referral Traffic

Referral traffic remains one of the clearest measurable outcomes from AI-assisted discovery.

Monitor:

• Sessions.
• Landing pages.
• Conversion rate.
• Revenue or lead value.
• Source platform.

Keep referral traffic separate from broader citation visibility because citations can create exposure without a click.

5. Track Branded Search Demand

Zero-click exposure can influence later branded searches.

Monitor:

• Branded impressions.
• Branded clicks.
• Brand-plus-product searches.
• Brand-plus-category searches.

A rise in branded demand can be a useful supporting signal of wider search visibility, although it should not automatically be attributed entirely to AI exposure.

6. Measure Direct Traffic Carefully

Users exposed to a brand inside an AI answer may later navigate directly.

Monitor changes in direct traffic, but treat them as supporting evidence rather than direct proof of AI attribution.

7. Segment by Query Type

AI Overview appearance varies by intent.

Report separately for:

• Informational.
• Question-based.
• Comparison.
• Commercial research.
• Transactional.
• Local.
• Branded.
• Non-branded.

This makes it easier to identify where generative visibility is growing fastest.

8. Segment by Content Type

Measure performance separately for:

• Research.
• Statistics pages.
• Service pages.
• Product pages.
• Guides.
• Comparison content.
• Video.
• Images.

Different content types may contribute to different stages of AI-assisted discovery.

9. Track Multi-Format Visibility

With Google’s multimodal reporting, businesses should monitor visibility generated through:

• Lens.
• Circle to Search.
• Image uploads.
• Visual product discovery.

This is especially important for industries where imagery directly influences selection.

10. Measure Commercial Outcomes

AI-search visibility should ultimately connect with business value wherever possible.

Track:

• Leads.
• Qualified leads.
• Transactions.
• Revenue.
• Pipeline.
• Customer value.

This prevents AI visibility from becoming a vanity metric.

The Measurement Rule

Google AI Overview Trends UK 2026–2027

The 50 statistics in this report point towards several clear changes in how Google Search is likely to develop through the remainder of 2026 and into 2027.

These are evidence-led trends rather than guaranteed outcomes.

They reflect current Google product development, UK AI adoption, AI Overview triggering patterns, citation behaviour, Search Console reporting and changing user expectations around generated answers.

1. AI Overviews Will Become a Normal Part of Search Measurement

AI Overview visibility is moving away from being something organisations estimate externally.

Google’s generative-search reporting now gives site owners increasingly direct insight into how content appears within AI-powered Search experiences.

Through 2026–2027, reporting is therefore likely to place more emphasis on:

• Generative impressions.
• Pages appearing in AI features.
• Device and country performance.
• Changes over time.

2. Ranking Reports Will Need an AI Layer

Because rankings and citation selection overlap imperfectly, conventional position tracking alone will provide an incomplete picture.

Search reporting will increasingly need to show:

• Organic ranking.
• Whether an AI Overview appears.
• Whether the brand is cited.
• Which page Google cites.

This will make generative visibility a second layer on top of traditional ranking analysis.

3. Informational Traffic Will Remain Under the Greatest Pressure

Question-based, explanatory and research-oriented searches currently show some of the highest AI Overview triggering rates.

These are also the searches where generated answers can most easily satisfy users without requiring an immediate click.

Publishers and businesses relying heavily on informational traffic should therefore expect continued pressure on conventional CTR.

4. Commercial Research Queries Will Become More Important

The expansion of AI Overviews into commercial-intent searches is strategically significant.

Through 2026–2027, businesses should pay particular attention to searches involving:

• Best providers.
• Best products.
• Alternatives.
• Comparisons.
• Recommendations.
• Pricing research.

These queries often influence the shortlist before the final transaction takes place.

5. Zero-Click Visibility Will Become a Bigger Measurement Problem

The evidence already shows that users are less likely to click conventional results when an AI summary appears.

This will increase pressure on businesses to measure:

• Branded search.
• Direct traffic.
• Assisted conversions.
• Citation visibility.
• Brand mentions.

Traffic will remain important, but it will represent a smaller portion of total search influence for some query types.

6. Citation Authority Will Become a More Established SEO Concept

The growing separation between ranking position and AI Overview source selection creates a distinct strategic question:

Why does Google trust one source enough to cite it?

This will increase interest in factors connected with:

• Original evidence.
• Institutional authority.
• Topical coverage.
• External references.
• Clear authorship.

7. Topic Architecture Will Matter More Than Isolated Keyword Targeting

Query fan-out means Google can explore several related searches behind one user prompt.

This rewards content architectures that cover connected subtopics rather than one isolated phrase.

Search strategy will therefore increasingly move from:

One Keyword → One Page

towards:

Topic → Subtopics → Evidence → Entity Relationships → Source Eligibility

8. Video Will Become More Important to Search Visibility

YouTube’s strong representation in AI Overview citation studies indicates that video is increasingly part of Google’s evidence ecosystem.

Businesses with genuine expertise will have stronger reasons to publish:

• Explainers.
• Research summaries.
• Product demonstrations.
• Technical walkthroughs.
• Industry analysis.

Transcripts and accurate descriptions will become increasingly valuable because they make video information easier to retrieve and understand.

9. Multimodal Search Will Move Into Mainstream SEO Reporting

Google’s September 2026 Search Console update introduced reporting for discovery through:

• Lens.
• Circle to Search.
• Image uploads.
• Chrome image search.

This is likely to increase the strategic value of high-quality visual assets.

For ecommerce, travel, property and other highly visual sectors, multimodal visibility may become a meaningful part of search reporting during 2027.

10. Natural-Language Search Will Continue to Expand

UK generative-AI adoption is already substantial, while Google reports significantly longer queries within AI Mode.

Users are becoming increasingly comfortable expressing complete questions rather than reducing their requirements to a few keywords.

This will favour pages that explain clearly:

• Who a service is for.
• What problem it solves.
• Its limitations.
• Its cost.
• Its alternatives.

11. Trust and Verification Will Become More Valuable

UK consumers are adopting generative AI while remaining concerned about accuracy.

That combination is likely to increase the value of:

• Primary sources.
• Research methodology.
• Original data.
• Transparent authorship.
• Clear citations.

Generative systems need sources, and users need reasons to trust the resulting answers.

12. SEO and GEO Will Continue to Converge

Google’s own guidance does not support the idea that AI Overview optimisation requires an entirely separate technical discipline.

Core foundations remain:

• Accessibility.
• Indexability.
• Useful content.
• Strong entities.
• Authority.
• Evidence.

Generative optimisation adds another objective: making those assets eligible for retrieval and citation.

SEO and GEO are therefore likely to become increasingly integrated rather than managed as disconnected programmes.

What This Means for UK Organisations

The likely direction is not that conventional SEO disappears.

It becomes one component of a broader visibility system.

Organisations will increasingly need to understand performance across:

• Organic rankings.
• AI Overview citations.
• AI Mode.
• Images.
• Video.
• Local discovery.
• Branded demand.

Research Methodology & Limitations

This report brings together 50 measurable data points relating to Google AI Overviews, AI-assisted search, source selection, click behaviour, UK AI adoption and the wider evolution of Google Search.

The objective is not to suggest that one statistic can describe the entire AI Overview environment.

Instead, the research combines multiple evidence types to show how generative search is changing visibility, click behaviour, citation selection and measurement.

1. Research Scope

The report examines five main evidence areas:

1. UK search and AI Overview adoption.
2. AI Overview triggering, query length and intent.
3. Click behaviour, zero-click search and traffic impact.
4. Sources, citations and AI Overview source selection.
5. AI Mode, multimodal search and wider UK AI adoption.

These areas were selected because they represent the main stages through which AI Overviews can affect search visibility and business performance.

2. Source Hierarchy

Sources were prioritised according to transparency, relevance and proximity to the underlying data.

The preferred evidence hierarchy was:

1. UK Government and regulator data — including Ofcom, ONS and DSIT.
2. First-party Google data — including product announcements, Search Console documentation and official Search guidance.
3. Independent behavioural research — including Pew Research Center.
4. Large-scale specialist SERP research — including Ahrefs and Semrush.
5. Transparent CGO Media calculations based on published source figures.

3. Geographic Classification

Not all meaningful AI Overview research is available as UK-only data.

To avoid presenting international or US findings as UK statistics, this report uses explicit evidence labels.

• UK Data — based on UK users, businesses or search activity.
• US Benchmark — evidence based specifically on US users or SERPs.
• International Benchmark — broader or non-UK search datasets.
• First-Party Platform Data — data published directly by Google.
• CGO Media Calculation — transparent arithmetic derived from published source data.

This distinction is especially important for:

• AI Overview prevalence.
• CTR impact.
• Citation behaviour.
• Source-selection studies.
• Commercial-search triggering.

4. Dataset Differences

Different sources measure different search populations.

For example:

• Ofcom may describe UK search behaviour.
• Pew may measure US browsing activity.
• Ahrefs may analyse millions of SERPs across a keyword database.
• Google may report global platform usage.

These datasets should not be compared as if they were collected from one uniform population.

5. AI Overview Appearance Rates Can Differ Between Studies

Different studies report different AI Overview prevalence rates.

This can result from differences in:

• Country.
• Device.
• Query set.
• Keyword intent.
• Measurement date.
• Personalisation.
• Search-interface changes.

For this reason, the report does not treat one AI Overview appearance percentage as a universal global rate.

6. Search Interfaces Change Rapidly

Google Search can change between the date a study is conducted and the date it is published.

Changes can include:

• AI Overview prevalence.
• Layout.
• Citation presentation.
• Expansion behaviour.
• AI Mode functionality.
• Search Console reporting.

This means even high-quality AI-search research has a shorter useful shelf life than many traditional SEO datasets.

7. CTR Studies Measure Association, Not Universal Causation

Studies showing lower click-through rates when AI Overviews appear do not establish that every individual website will experience the same reduction.

CTR can vary by:

• Ranking position.
• Query type.
• Brand recognition.
• SERP layout.
• Device.
• User intent.
• AI Overview size.

CTR findings should therefore be treated as benchmark evidence rather than guaranteed traffic-loss percentages.

8. Zero-Click Does Not Automatically Mean Zero Value

A search may end without a website click while still influencing a later decision.

A user can:

• Remember a brand.
• Search the brand later.
• Navigate directly.
• Purchase through another channel.
• Use another search or AI system.

Current analytics systems cannot always reconstruct that complete journey.

Zero-click exposure should therefore not automatically be classified as either commercially valuable or commercially worthless.

9. Citation Does Not Equal Traffic

Being cited inside an AI Overview does not guarantee that the user will click the source.

Pew’s behavioural research shows that clicks on AI-cited sources can be rare.

Citation visibility should therefore be measured separately from:

• Referral traffic.
• Conversions.
• Revenue.

10. Citation Frequency Does Not Automatically Prove Authority

A heavily cited domain can appear frequently because it contains:

• Large amounts of content.
• User-generated discussion.
• Video transcripts.
• Highly indexed information.

Citation frequency alone should not be interpreted as proof that one domain is more authoritative in every subject area.

11. Source-Selection Research Has Visibility Limits

External researchers can observe citations and SERPs but do not have access to Google’s complete internal retrieval and ranking systems.

As a result, studies can identify relationships and patterns but cannot prove the full weighting of Google’s internal source-selection criteria.

12. Query Fan-Out Is Only Partly Observable

Google confirms that AI-powered Search can use query fan-out.

However, external researchers do not have complete visibility into every subquery Google generates for every user request.

This means source-selection analysis can explain part of the retrieval process without revealing the entire internal sequence.

13. First-Party Google Data Requires Context

Google’s own usage figures are useful for understanding the scale and direction of its products.

However, platform-published metrics should still be interpreted according to the definition provided.

Examples include:

• Monthly active users.
• Available devices.
• Query length.
• Feature rollout.

These measures are not directly interchangeable with market share, search volume or UK adoption.

14. AI Mode and AI Overviews Are Different Products

AI Mode and AI Overviews both use generative technology but should not be treated as the same search interface.

AI Overviews typically appear within conventional results pages.

AI Mode provides a more conversational, exploratory search experience.

Usage data for one product should therefore not be automatically attributed to the other.

15. Business AI Adoption Does Not Equal AI Search Usage

ONS and DSIT data showing business AI adoption provide context for the wider technological environment.

They do not measure how frequently those businesses use Google AI Overviews specifically.

Business AI adoption is included because it demonstrates the broader normalisation of generative technologies across the UK economy.

16. Consumer Generative-AI Use Does Not Equal Google AI Overview Use

Similarly, a consumer who uses ChatGPT, Gemini or another generative tool is not necessarily an AI Overview user.

These figures provide context for changing consumer expectations around generated information, not direct AI Overview adoption rates.

17. Commercial-Intent Classifications Can Differ

Search-research platforms classify intent algorithmically.

A query may be classified as:

• Informational.
• Commercial.
• Transactional.
• Local.

Different providers may classify the same query differently.

Intent-based AI Overview appearance rates should therefore be interpreted according to the methodology of the individual study.

18. Correlation Does Not Establish a Ranking Factor

If commonly cited pages share characteristics such as high authority, original research or strong conventional rankings, that does not prove Google directly uses one of those characteristics as a standalone AI Overview ranking factor.

The report therefore avoids converting observed correlations into unsupported algorithm claims.

19. There Is No Verified Universal “AI Overview Optimisation Formula”

No public evidence establishes a guaranteed combination of factors that will cause Google to cite a page in an AI Overview.

Claims promising guaranteed AI citations should therefore be treated cautiously.

Google’s current public guidance continues to emphasise established Search foundations rather than a separate technical AI Overview checklist.

20. This Research Will Require Regular Updating

AI search is developing rapidly.

This page should be updated when:

• New official UK data becomes available.
• Google changes AI Overview or AI Mode functionality.
• Search Console adds new reporting.
• New clickstream studies improve behavioural evidence.
• Large citation studies update source-selection patterns.
• Older benchmarks become unrepresentative.

Research Updates

CGO Media treats this page as a living research resource.

Older statistics may be replaced when stronger evidence becomes available, particularly where:

• A larger dataset is published.
• Official UK evidence becomes available.
• A Google product changes materially.
• A methodology becomes outdated.

For the wider CGO Media research approach, see the CGO Media Research Methodology.

Frequently Asked Questions — Google AI Overview Statistics UK 2026

The following questions address some of the most common issues surrounding Google AI Overviews, citation visibility, organic traffic, AI Mode and generative search optimisation.

What is a Google AI Overview?

A Google AI Overview is a generated summary that can appear within Google Search in response to certain queries.

It can combine information from multiple sources and display supporting links or citations alongside the generated answer.

The feature is designed to help users understand a topic without requiring them to open every individual result.

How common are AI Overviews in Google Search?

There is no single universal appearance rate.

Prevalence varies according to:

• Country.
• Device.
• Query type.
• Search intent.
• Measurement date.
• Dataset methodology.

Ofcom reports that around 30% of Google searches show AI Overviews within the UK search environment, while large third-party SERP studies have measured different rates across their own datasets.

Which searches are most likely to trigger AI Overviews?

Current research shows particularly high appearance rates for:

• Informational searches.
• Question-based searches.
• Longer queries.
• Reasoning queries.
• Definition queries.
• Explanatory searches.

Local and highly transactional queries currently tend to trigger AI Overviews less frequently than explanatory searches.

Do longer queries trigger more AI Overviews?

Yes, current benchmark research shows a clear relationship between query length and AI Overview appearance.

Longer searches provide more context about what the user wants, making them particularly suitable for generated synthesis.

This is important because AI Mode also encourages substantially longer and more conversational queries than conventional search.

Do Google AI Overviews reduce organic clicks?

Independent studies indicate that conventional-result click-through rates can be lower when an AI Overview appears.

However, there is no universal percentage traffic loss that applies to every website.

Impact depends on:

• Query intent.
• Organic ranking position.
• AI Overview length.
• SERP layout.
• Brand recognition.
• Whether the generated answer satisfies the user’s need.

Can a page still rank first but lose traffic because of an AI Overview?

Yes.

A page can retain the same conventional ranking position while receiving fewer clicks if an AI Overview answers more of the user’s question before the organic result is reached.

This is why ranking and traffic now need to be measured separately.

Does being cited in an AI Overview generate traffic?

Sometimes, but not necessarily.

Citation visibility can create:

• Brand exposure.
• Source attribution.
• Authority signals.
• Future branded searches.

But independent clickstream research shows that users may not frequently click the cited source directly.

Citation visibility and referral traffic should therefore be measured separately.

Does a page need to rank in Google’s top ten to be cited?

No.

Large citation studies show significant overlap between high-ranking pages and AI Overview sources, but many cited URLs appear below the conventional top ten or even outside the top 100 for the original query.

Google can use query fan-out to explore related searches and retrieve supporting evidence from other result sets.

What is query fan-out?

Query fan-out is the process through which Google can issue multiple related searches or subqueries behind the scenes when generating an AI-powered response.

This allows Google to gather information around:

• Subtopics.
• Comparisons.
• Definitions.
• Supporting evidence.
• Related entities.

For SEO, this means topical depth can matter more than targeting only one exact keyword.

What types of websites does Google cite in AI Overviews?

Citation studies show a diverse source ecosystem.

Frequently observed source types include:

• Traditional publishers.
• Specialist websites.
• YouTube.
• Reddit.
• Wikipedia.
• Government websites.
• Community platforms.

This demonstrates that AI Overview source selection is not limited to conventional editorial websites.

Why is YouTube important for AI Overview visibility?

Recent citation research has found YouTube among the most frequently cited domains in AI Overviews.

Google can potentially use:

• Video titles.
• Descriptions.
• Transcripts.
• Related metadata.

This means video increasingly forms part of the wider search evidence ecosystem.

Is Reddit important for Google AI Overviews?

Reddit appears prominently in several large AI Overview citation datasets.

Community discussions can provide:

• Real-world experience.
• Product opinions.
• Comparisons.
• Problem-solving discussions.

This does not mean businesses should manufacture Reddit activity.

It demonstrates that Google can draw from user-generated evidence alongside conventional publisher sources.

Is there special schema markup for Google AI Overviews?

No verified special “AI Overview schema” exists.

Businesses should use supported structured data only where it accurately describes visible content.

Relevant structured data may include:

• Article.
• Organization.
• Dataset.
• Product.
• LocalBusiness.
• VideoObject.

Structured data can help search engines understand content, but it does not guarantee AI Overview inclusion.

Can businesses optimise specifically for AI Overviews?

Businesses can improve their eligibility, but there is no verified formula that guarantees a citation.

The strongest current approach is to improve:

• Technical search accessibility.
• Topical coverage.
• Original evidence.
• Organisational clarity.
• External authority.
• Multi-format content.

These foundations support both conventional SEO and generative search visibility.

Do AI Overviews replace SEO?

No.

AI Overviews operate within the Google Search ecosystem and still depend on web content and searchable sources.

The emerging change is that businesses increasingly need to compete for both:

• Organic ranking visibility.
• Generative citation visibility.

What is the difference between AI Overviews and AI Mode?

AI Overviews generally appear as generated summaries inside conventional Google Search results.

AI Mode is a more conversational search experience designed for longer, more complex questions and follow-up exploration.

The products are related but should not be treated as identical.

How should businesses measure AI Overview visibility?

A useful measurement system should include:

• Generative-search impressions.
• AI Overview appearance rate.
• Citation frequency.
• Organic ranking position.
• AI referral traffic.
• Branded-search change.
• Leads and revenue.

Citation visibility should not be treated as the final commercial metric.

Can AI Overview visibility influence a business without generating a click?

Yes.

A user may see the organisation named or cited and later:

• Search for the brand.
• Visit directly.
• Compare the company elsewhere.
• Return during a later buying stage.

This creates a wider attribution problem because not every generative interaction creates an immediate referral.

Should businesses publish original research for AI search?

Original research can be particularly valuable because it makes the organisation a primary source rather than another publisher repeating existing information.

Useful assets include:

• Datasets.
• Statistics.
• Surveys.
• Testing.
• Case studies.
• Original analysis.

Original research does not guarantee citation, but it can create stronger source differentiation.

Will AI Overviews become more commercial?

Recent benchmark research indicates AI Overviews are appearing more frequently on commercial research queries.

This makes comparison, recommendation and provider-selection searches particularly important to monitor.

However, immediate transactional and local searches currently show different triggering patterns and should not be assumed to evolve identically.

What is the biggest mistake businesses can make with AI Overview optimisation?

One of the biggest mistakes is chasing supposed AI-search tricks while neglecting search fundamentals.

A stronger priority order is:

Conclusion — Final Research Findings from the 50 Google AI Overview Statistics UK 2026

The 50 statistics in this report show that Google AI Overviews are changing both the structure of search results and the relationship between visibility, citations and website traffic.

The evidence does not support the idea that traditional SEO is disappearing.

Instead, conventional organic rankings are becoming part of a broader search environment that now includes generated answers, cited sources, conversational search, images, video and multimodal discovery.

AI Overviews Are Now Part of Mainstream Search

Google already operates inside an enormous UK search market.

AI Overviews are now appearing across a meaningful proportion of searches and are encountered by a substantial share of UK adults.

This makes generative visibility an operational search issue rather than an experimental technology question.

Informational Search Is Being Restructured First

The strongest AI Overview triggering patterns currently appear around:

• Questions.
• Explanations.
• Definitions.
• Reasoning.
• Longer informational searches.

These are precisely the searches that have historically generated large amounts of organic informational traffic.

The result is a structural shift in how upper-funnel visibility can translate into website visits.

Commercial Search Is Increasingly Entering the Generative Layer

The newer data suggests AI Overviews are moving beyond purely informational searches into commercial research.

That makes searches involving:

• Comparisons.
• Best providers.
• Best products.
• Alternatives.
• Recommendations.

increasingly important.

These searches can influence which businesses enter the customer’s consideration set before a direct transaction takes place.

Organic Rankings Still Matter — But They No Longer Explain Everything

The citation evidence shows substantial overlap between conventional rankings and AI Overview source selection.

However, that overlap is incomplete.

Google can cite pages that:

• Rank below the first page.
• Appear within other SERP features.
• Do not rank in the top 100 for the original query.
• Surface through related query fan-out.

This means conventional ranking authority remains important, but it is not the only route into Google’s generated answers.

Citation Visibility Is Becoming a Separate Search Asset

A business can increasingly gain search exposure by being cited within the generative layer even where the conventional ranking position is weaker.

This creates a second form of visibility:

The two should increasingly be measured separately.

Visibility and Traffic Are Separating

Independent behavioural research shows that users are less likely to click traditional search results when an AI-generated summary appears.

The same research also suggests that citations themselves do not automatically produce significant referral traffic.

This creates a new measurement challenge:

Zero-Click Search Is Becoming More Sophisticated

Zero-click behaviour existed long before AI Overviews.

Generated summaries extend that behaviour into more complex searches by allowing Google to synthesise information that users previously needed several websites to gather.

Businesses therefore need to consider whether search visibility is influencing:

• Brand awareness.
• Branded searches.
• Direct visits.
• Future supplier consideration.
• Later conversions.

even where no immediate organic referral occurs.

Query Fan-Out Changes What It Means to Optimise for a Search

One user query can now lead Google to explore several related questions and subtopics.

That shifts the strategic target from:

Rank One Page for One Keyword

towards:

Build Authority Across the Entire Information Space Around the Query

This makes knowledge architecture, topic coverage and internal linking increasingly important.

Source Type Matters

AI Overview citation data shows substantial visibility for:

• YouTube.
• Reddit.
• Government websites.
• Wikipedia.
• Social platforms.
• Specialist publishers.

Google’s generated answers therefore draw from a much wider information ecosystem than conventional commercial websites alone.

Original Evidence Becomes More Strategically Valuable

If generative systems need evidence to construct answers, organisations producing original evidence have an opportunity to become primary sources.

Research assets such as:

• Statistics.
• Datasets.
• Surveys.
• Experiments.
• Case studies.
• Frameworks.

can potentially create value beyond their direct organic traffic.

They can contribute to citation visibility, digital PR, external references and wider entity authority.

Trust Is Becoming More Important, Not Less

UK consumers are adopting generative AI rapidly while remaining concerned about inaccurate AI-generated information.

This increases the strategic importance of:

• Transparent methodology.
• Named sources.
• Clear authorship.
• Research provenance.
• Organisational accountability.

Search systems need reliable evidence, and users need confidence in the answers they receive.

Search Is Becoming Multimodal

Google’s expansion into AI Mode, Lens, Circle to Search and multimodal Search Console reporting shows that search is moving beyond text queries.

Search visibility increasingly involves:

• Text.
• Images.
• Video.
• Voice.
• Structured product and entity information.

For many sectors, SEO strategy will need to account for all of these formats.

SEO and GEO Are Converging

The research does not support abandoning traditional SEO for a completely separate generative optimisation discipline.

The strongest foundations remain familiar:

• Technical accessibility.
• Useful content.
• Clear information architecture.
• Strong entities.
• Authority.
• Trust.

Generative search adds another objective on top of those foundations:

Become Eligible to Be Retrieved and Cited as Evidence

Final CGO Media Research Position

The most important change is not simply that Google now generates answers.

It is that the search engine increasingly decides:

• Which questions need synthesis.
• Which related subqueries should be explored.
• Which sources should contribute.
• Which evidence should be displayed.

That moves search optimisation beyond competition for one ranked link.

Research Usage, Citation & Press

The Google AI Overview Statistics UK 2026 research is intended to support journalists, researchers, businesses, agencies, academics and organisations examining the development of generative search and its impact on digital visibility.

Statistics, findings and CGO Media analysis from this report may be referenced provided the original evidence classification and relevant limitations are preserved.

How to Cite This Research

A suggested citation format is:

Where a specific statistic is quoted, readers should also consult and reference the original source identified beneath that statistic where appropriate.

Journalists & Media

Journalists may quote individual statistics, findings or CGO Media analysis from this report in editorial coverage.

Where possible, please:

• Credit CGO Media.
• Link to the original research page.
• Preserve the statistic’s geographic classification.
• Distinguish measured source data from CGO Media interpretation.
• Retain significant methodological caveats.
• Avoid presenting benchmark findings as guaranteed outcomes for every website.

For press enquiries, methodology questions, research commentary or requests for supporting material, visit:

CGO Media Press & Media

Researchers & Academics

Researchers may reference the dataset provided that the different evidence types are preserved.

This report is not based on one single survey or experiment.

It combines:

• UK regulatory evidence.
• UK Government statistics.
• First-party Google platform data.
• Independent browsing research.
• Large-scale SERP analysis.
• Transparent CGO Media calculations.

The 50 statistics should therefore be interpreted as an evidence synthesis rather than one statistically homogeneous sample.

Businesses & Organisations

Businesses may use the findings for:

• AI search strategy.
• SEO planning.
• GEO planning.
• Board presentations.
• Search investment cases.
• Content strategy.
• Digital PR planning.
• AI visibility reporting.

External benchmark data should not replace the organisation’s own first-party performance data.

The strongest business analysis combines external research with:

• Google Search Console.
• Analytics.
• CRM data.
• AI referral traffic.
• Branded-search trends.
• Leads.
• Revenue.

Using Individual Statistics

When reproducing an individual statistic, retain its evidence classification.

The second statement would incorrectly convert an international search benchmark into a UK national statistic.

Using Traffic-Impact Statistics

CTR and traffic-impact findings require particular care.

A measured reduction in one study should not be presented as a guaranteed traffic loss for every website.

A responsible citation might state:

This preserves the study context instead of converting the result into a universal prediction.

Using Source & Citation Research

Domain citation shares should also be quoted with their measurement context.

For example, a domain’s “mention share” among the top 50 cited sources is not the same as its percentage of every AI Overview citation globally.

The distinction should remain clear when reporting source-selection data.

Using CGO Media Calculations

Where CGO Media performs arithmetic using published source figures, the result is labelled CGO Media Calculation.

These calculations may be cited where the underlying source values remain identifiable.

Example:

Charts, Infographics & Visualisations

Charts, figures and visual summaries created from this research may be referenced with attribution to:

CGO Media Research Team — Google AI Overview Statistics UK 2026

Where a visual reproduces third-party data, the underlying source should also remain identifiable.

Research Corrections & Updates

AI-assisted search changes rapidly.

CGO Media may update this report when:

• Google changes AI Overview functionality.
• New Search Console reporting becomes available.
• New UK evidence is published.
• Stronger clickstream data becomes available.
• Large citation studies are updated.
• An existing source publishes revised figures.
• A material error or ambiguity is identified.

Older statistics may be replaced rather than retained solely for historical continuity when newer evidence provides a more accurate picture of the current search environment.

Research Methodology

For a broader explanation of CGO Media’s source selection, evidence classification, calculation standards and research interpretation, visit:

CGO Media Research Methodology

Sources & References

The 50 statistics in this research are drawn from UK regulatory and government evidence, first-party Google data, independent behavioural research and large-scale search-industry datasets.

CGO Media has prioritised original sources wherever possible rather than relying on secondary articles reporting another organisation’s findings.

1. Ofcom — Online Nation / UK Search Behaviour

Ofcom. From Apps to AI Search: How the UK Goes Online in 2025. Published 10 December 2025.

Used for UK evidence including:

• 82% of UK adults using Google Search.
• Approximately 3 billion UK Google searches per month.
• Around 30% of searches displaying AI Overviews.
• 53% of adults saying they often see AI-generated search summaries.
• 1.8 billion UK ChatGPT visits from January to August 2025.
• 368 million corresponding visits during January to August 2024.

View Ofcom source

2. Google — AI Search Scale & Website Owner Controls

Google. New Opportunities, Control and Insights for Website Owners. Published 3 June 2026; updated 31 August 2026.

Used for first-party platform evidence including:

• More than 2.5 billion monthly active users for AI Overviews.
• More than 1 billion monthly users for AI Mode.
• The increasing role of generative AI within Google Search.

View Google source

3. Google Search Console — Generative AI Performance Reporting

Google Search Central. Introducing Search Generative AI Performance Reports in Search Console. Published 3 June 2026.

Google states that the dedicated generative-AI performance reports were made available to all websites worldwide by 31 August 2026.

The reports include:

• Generative-search impressions.
• Pages appearing within AI features.
• Countries.
• Devices.
• Performance over time.

View Google Search Console source

4. Ahrefs — AI Overview Triggering Study

Ryan Law, Ahrefs. What Triggers AI Overviews? 86 Factors and 146 Million SERPs Analyzed. Published 10 November 2025.

The study analysed approximately 146 million desktop SERPs measured in September 2025.

Used for evidence including:

• Overall AI Overview prevalence within the Ahrefs dataset.
• Informational-intent association.
• Branded versus non-branded triggering.
• Query-length differences.
• Question-query triggering.
• Reason, definition, instruction and comparison queries.
• Local-search triggering.

View Ahrefs triggering study

5. Semrush — Commercial-Intent AI Overview Study

Luke Harsel, Semrush. AI Overviews Are Expanding Across Commercial Intent Search. Published 2 July 2026.

The study analysed more than 600,000 keywords across 10 industries over six months.

Used for evidence including:

• 71% growth in AI Overview appearance on commercial-intent SERPs.
• A 5% decline on transactional-intent SERPs.
• 231% growth for commercial finance searches within the study.

View Semrush study

6. Pew Research Center — AI Summary Click Behaviour

Pew Research Center. Google Users Are Less Likely to Click on Links When an AI Summary Appears in the Results. Published 22 July 2025.

The study analysed browsing behaviour from 900 US adults and tens of thousands of Google searches conducted during March 2025.

Used for evidence including:

• Traditional search-result clicks of 8% when an AI summary appeared versus 15% without one.
• Approximately 1% of AI-summary visits generating a click on a cited source.
• 26% of AI-summary visits ending the browsing session versus 16% without an AI summary.
• Approximately two-thirds of searches producing no click on a listed result.
• Source-distribution evidence involving Wikipedia, YouTube, Reddit and government websites.

View Pew Research Center study

7. Ahrefs — AI Overview CTR Impact

Ryan Law, Ahrefs. Update: AI Overviews Reduce Clicks by 58%. Published 4 February 2026.

The updated analysis compared approximately:

• 150,000 informational keywords with AI Overviews.
• 150,000 informational keywords without AI Overviews.

The study used aggregated Google Search Console data to estimate CTR differences.

Used for the report’s benchmark evidence concerning reduced organic CTR when AI Overviews appear.

View Ahrefs CTR study

8. Ahrefs — AI Overview Citation & Ranking Overlap

Louise Linehan, Ahrefs. Update: 38% of AI Overview Citations Pull From the Top 10. Published 2 March 2026.

This research examined large-scale AI Overview citation data to compare cited pages with their conventional Google rankings.

Used for evidence including:

• Approximately 38% of cited URLs appearing within the first ten result blocks.
• Substantial citation activity below the top ten.
• AI Overview citations coming from URLs outside the top 100.
• The role of YouTube among cited URLs.

View Ahrefs citation study

9. Ahrefs — Most-Cited AI Overview Domains

Si Quan Ong, Ahrefs. The 50 Most-Cited Websites in Google AI Overviews. September 2026 dataset.

The US dataset measures mention share among leading AI Overview sources.

Used for evidence concerning highly cited domains including:

• YouTube.
• Reddit.
• Facebook.
• Google.
• Instagram.
• Quora.
• Wikipedia.

Mention share in this dataset should not be interpreted as a domain’s share of every Google AI Overview citation globally.

View Ahrefs citation-domain data

10. Google — AI Mode Launch in the UK

Google. Google Search: Introducing AI Mode in the UK. Published 28 July 2025.

Used for evidence and context including:

• UK rollout of AI Mode.
• Early AI Mode queries being two to three times longer than traditional searches.
• Use of query fan-out.
• AI Mode’s focus on complex, multi-part and exploratory searches.

View Google UK AI Mode source

11. Alphabet / Google — AI Mode Query Behaviour

Alphabet / Google. Alphabet Earnings, Q4 2025: CEO’s Remarks.

Used for first-party evidence including:

• AI Mode queries being approximately three times longer than traditional searches.
• Nearly one in six AI Mode queries using voice or images.
• Circle to Search availability on more than 580 million Android devices.

View Alphabet / Google source

12. DSIT — UK Public Engagement with Artificial Intelligence

Department for Science, Innovation and Technology. DSIT Public Engagement Survey 2025/2026.

Used for UK evidence including:

• 59% of UK adults using some form of generative AI during the previous three months.
• 56% using AI capable of producing human-like text or speech.
• 54% using AI-powered digital assistants capable of understanding natural language.
• 74% concerned that AI-generated information may be inaccurate.
• 47% identifying easier access to information or advice as an AI benefit.

View DSIT survey

13. Office for National Statistics — AI in UK Businesses

Office for National Statistics. Artificial Intelligence in UK Businesses: 2023 to 2026. Released 20 July 2026.

Used for UK business evidence including:

• AI use among businesses with 10 or more employees increasing from approximately 12% in late 2023 to approximately 35% in June 2026.
• 49% adoption among businesses with 250 or more employees.

View ONS business AI research

14. Office for National Statistics — UK AI Thematic Account

Office for National Statistics. Measuring Artificial Intelligence in the UK Economy Using a Thematic Account. September 2026.

Used for evidence including:

• 17% of UK businesses with 10 or more employees using large-language-model text generation in June 2026.
• 14% using visual-content creation.
• The increase in LLM text-generation adoption since September 2023.

View ONS thematic account

15. Google Search Console — Multimodal Search Reporting

Google Search Central. Announcing Web Multimodal Search Performance Reporting in Search Console. Published 24 September 2026.

Used for current search-measurement context covering:

• Google Lens.
• Circle to Search.
• Image uploads to Google Search.
• Chrome image-search functionality.

View Google multimodal reporting source

Supporting Google Documentation

The following first-party Google resources were also used when interpreting optimisation, measurement and source-selection implications.

Google’s Guide to Optimising for Generative AI Features

Google’s current guidance explains that established Search fundamentals continue to apply to generative AI Search features and provides guidance on measuring visibility through Search Console.

View Google AI Search guidance

Preferred Sources in AI Search

Google announced in May 2026 that Preferred Sources were being brought into AI Overviews and AI Mode.

View Google Preferred Sources announcement

Research Review Date

This source list was reviewed against available evidence on 28 September 2026.

Because Google AI Search is developing rapidly, CGO Media may replace individual statistics when newer primary evidence, larger datasets or more representative UK research becomes available.