AI Overviews launched broadly in the U.S. in May 2024, then expanded internationally; by May 20, 2025, Google said AI Overviews were available in 200+ countries and territories and 40+ languages.
For UK marketing teams, this is the biggest shift in organic search dynamics in years. The uncomfortable reality: when AIO appears, fewer users click through to websites even if you rank #1.
The Traffic Collapse: Real Numbers From Real Publishers
Let’s be clear: the impact varies by query type and vertical, but multiple independent datasets point to the same direction of travel.
Users click less when AI Overviews appear
The Pew Research Center analysed real-world browsing behaviour and found users were less likely to click when an AI summary appeared: clicks to traditional results happened in 8% of visits with an AI summary vs 15% without.
If your SEO strategy is built on capturing informational clicks (“what is…”, “how to…”, “best…”) this trend hits your pipeline directly.
The #1 result loses clicks even if rankings don’t move
Ahrefs analysed a large keyword set and reported that the presence of an AI Overview correlated with a 34.5% lower average clickthrough rate (CTR) for the top-ranking page compared to similar informational keywords without AIO.
That’s the key nuance: you can “hold position” and still lose traffic because the interface has changed.
The “89% drop” claim: where it comes from (and why it matters)
The most dramatic headline figure, up to an 89% drop comes from reporting on evidence submitted by DMG Media to the UK’s Competition and Markets Authority, claiming desktop visits fell from 25.2% of users clicking to 2.8% when AI Overviews appeared (Campaign’s coverage of the DMG evidence).
That isn’t the average outcome across all sites and queries. It’s a worst-case scenario, but it’s also the scenario that matters most if:
- you’re a publisher, aggregator, or ad-monetised content business
- your funnel depends on high-volume informational queries
- your content is “summarisable” (definitions, basic how-tos, simple lists)
One Exception: Why Branded Searches Are Thriving
There’s one pattern that can work in your favour: brand.Amsive’s research found that branded keywords that triggered AI Overviews saw CTR increase by +18.68% on average.
In plain English: when people already recognise or trust a name, AIO can reinforce authority and drive “validation clicks” for depth, proof, or next steps.
This is the emerging split:
- generic content gets summarised
- brands get remembered
Google's Counter-Narrative: More Features, More Partnerships
Google argues that AI Overviews create “new opportunities for discovery” and has been rolling out features that may help publishers, but they don’t remove the core issue (AIO absorbs demand on-page).
Preferred Sources goes global
In December 2025, Google expanded Preferred Sources globally for English-language users, letting users “star” outlets they trust so those sources appear more often in Top Stories.
Subscriptions get spotlighted
Google also announced Spotlighting subscriptions, which highlights links from users’ paid news subscriptions in Gemini and said it would appear in Search surfaces including AI Overviews / AI Mode.
Search Console Insights is experimenting with social channel reporting
Google has tested adding social channels into Search Console Insights so site owners can see performance across web + social surfaces (Search Engine Land coverage and Google’s help documentation).
AI Mode changes: more inline links + “why this link” context
Google has said it is increasing inline links in AI Mode and adding context on why a link may be useful (Search Engine Land coverage).
What This Means for UK Marketing Teams
If your business relies on organic search traffic, this affects you immediately.
- Your content marketing ROI is changing
That blog post ranking #1 used to drive predictable traffic. Now it may be cited in an AI Overview while CTR drops materially. - Top-of-funnel acquisition is becoming more volatile
Informational queries are most exposed. - Brand becomes a defensive moat
If branded queries can benefit while generic ones lose, brand-building becomes traffic protection. - Competitors are still measuring rankings, not outcomes
Many teams haven’t updated measurement and decision-making for “AIO-first SERPs.”
The Legal Landscape: Why Publishers Are Pushing Back
This is no longer just an SEO debate, it’s an economics debate.
Chegg vs Google
Chegg sued Google in February 2025 (Search Engine Land coverage).
Penske Media vs Google
Penske Media (Rolling Stone / Variety) sued Google in September 2025 (TechCrunch coverage).
UK pressure and CMA attention
UK publishers and industry groups have been pushing for stronger controls and scrutiny around the impact of AI summaries on media economics (Campaign’s reporting).
How to Control (or Reduce) Your Content in AI Overviews
Not every business wants to be summarised, especially if the page exists to drive demos, leads, subscriptions, or other high-value conversions.
Use snippet controls to limit what can be shown in Search features
These controls affect what Google can extract and display as snippets (including in many SERP features).
Option 1: nosnippet (blunt instrument)
<meta name="robots" content="nosnippet">
Option 2: data-nosnippet (surgical control - recommended)
<div data-nosnippet>
This section won’t be used in snippets.
</div>
Option 3: max-snippet (limit snippet length)
<meta name="robots" content="max-snippet:50">
Use Google-Extended for Gemini training/grounding
Google-Extended controls whether your content can be used for training future Gemini models and for grounding Gemini experiences — but it does not block Google from using your content in AI Overviews (explained here).
Example:
User-agent: Google-Extended
Disallow: /premium-content/
Bottom line:
- Want to limit extraction in Search → use snippet controls (
data-nosnippet,nosnippet,max-snippet) - Want to limit Gemini training/grounding → use
Google-Extended
What UK Marketing Teams Should Do Now
Phase 1: Immediate (Week 1–2) - quantify exposure
- Pick your top 50 queries and manually log: AIO present? (Y/N)
- Segment GSC performance by intent (informational vs commercial vs branded).
- Identify pages most likely to be summarised (definitions, basic how-tos, listicles).
- Protect high-conversion sections with
data-nosnippet(pricing explanations, unique frameworks, premium research, demo/pitch logic).
Phase 2: Weeks 3–6 - restructure around topical authority
AI Overviews pull from a topic area, not a single URL.
- Build topic clusters (pillar + supporting pages) with strong internal linking.
- Put a concise answer early, then add depth AI can’t fully replace (UK-specific nuance, examples, templates, original data).
- Publish “angle coverage” for each pillar: definition, how it works, mistakes, comparisons, implementation, costs, and measurement.
Phase 3: Weeks 7–10 - technical + trust signals
- Add appropriate schema only where it’s genuine (FAQ where real FAQs exist, HowTo where it’s truly a process).
- Strengthen author credibility (bio, credentials, proof of experience, editorial review).
- Add “last reviewed / updated” dates where you actually maintain content.
- Improve assets (diagrams, calculators, downloadable checklists/templates).
Phase 4: Months 3–6 - traffic diversification (non-negotiable)
If Google is less predictable, your growth can’t rely on one channel.
- Email capture (content upgrades, templates, tools)
- YouTube (still heavily surfaced in Google ecosystems)
- LinkedIn (especially for UK B2B)
- Community presence (forums/industry spaces)
- Proprietary tools and datasets (things AIO can’t replace)
Measuring Success in the AI Overview Era
Rankings alone won’t tell the truth anymore.
Primary KPIs
- AIO presence rate (for your tracked query set): % of priority queries showing AIO
- Citation share (manual / tooling): how often you’re cited when AIO appears
- Engagement quality: time on page, pages per session, conversion rate for users who click through
A note on attribution tracking
Avoid hard claims like “Google always appends X UTM parameter.” Instead:
- monitor the pages most frequently cited
- compare performance trends as AIO presence increases
- use controlled query sets + SERP snapshots as your ground truth
The Bottom Line: Adapt or Get Quietly Replaced
AI Overviews aren’t a temporary experiment. They’re a new interface layer between you and your customers.
If your strategy is still “rank #1 and collect clicks,” you’re playing the old game. The new game is:
- Be the brand that gets cited and remembered
- Publish depth AI can’t fully compress
- Build assets that require interaction (tools, templates, calculators, data)
- Capture demand through owned channels so you’re not hostage to SERP layouts
Google AI Overviews FAQs
What are Google AI Overviews?
AI Overviews are Google’s AI-generated summaries that appear at the top of some search results. They synthesise information from multiple sources and may include citation links to publisher pages for deeper reading.
When did AI Overviews launch in the UK?
Google launched AI Overviews broadly in the U.S. in May 2024 and expanded internationally. Google stated AI Overviews were available in 200+ countries and territories and 40+ languages by May 20, 2025.
Do AI Overviews reduce organic traffic?
In many cases, yes. Research indicates users click less often when an AI summary appears. Pew found clicks to traditional results occurred in 8% of visits with an AI summary vs 15% without.
How big can the traffic drop be?
It varies by query type and SERP layout. In worst-hit examples, reporting on DMG Media evidence suggested desktop click-through fell from 25.2% to 2.8% when AI Overviews appeared (about an 89% drop).
Do AI Overviews reduce clicks even if you rank #1?
Often, yes. Ahrefs reported a 34.5% lower average CTR for the top-ranking page on informational keywords when AI Overviews were present versus similar keywords without AI Overviews.
Are all keywords impacted equally?
No. Informational queries (“what is”, “how to”, “best”) are generally more exposed to AI Overviews than branded or highly transactional queries. Your impact depends on intent, vertical, and whether your content can be summarised on-SERP.
Can AI Overviews ever increase traffic?
They can for recognised brands and high-trust sources. Amsive found branded keywords triggering AI Overviews saw CTR increase by +18.68% on average.
How do we measure AI Overview impact for our site?
Start with a controlled approach:
- track a fixed set of priority queries and log whether AI Overviews appear
- monitor CTR and clicks for those queries over time in Search Console
- compare engagement on pages that are frequently cited vs those that aren’t
Avoid relying on “universal” UTM parameter assumptions—treat SERP snapshots and query sets as your ground truth.
Can we block our content from appearing in AI Overviews?
You can limit what Google can extract for snippets using controls like nosnippet, data-nosnippet, and max-snippet. These can reduce how much text is eligible to appear in SERP snippets and some features.
Does Google-Extended block AI Overviews?
No. Google-Extended is generally understood as a control for Gemini training/grounding usage, not a way to block inclusion in AI Overviews in Search.
Should we block AI Overviews for high-conversion pages?
Consider selectively limiting extraction (e.g., using data-nosnippet on key sections) for pages where you need the click to convert—like pricing logic, proprietary frameworks, gated research, or demo-driving content. For awareness content, being cited can still be valuable.
What should UK marketing teams do first?
A fast first step is to:
- pick your top 50 queries and log where AI Overviews appear
- segment GSC performance by intent (informational vs commercial vs branded)
- identify which pages are most exposed (definitions/how-tos/listicles)
- protect high-conversion sections with
data-nosnippet
Are publishers taking legal action against Google?
Yes—there have been high-profile lawsuits and ongoing scrutiny. For example, Chegg sued Google in February 2025. Search Engine Land coverage Penske Media sued Google in September 2025.
