Google AI Overviews now appear at the top of millions of searches. This guide explains how AI Overviews work, what determines which sources get cited, and the optimisation tactics that increase your visibility.
TL;DR
- Google AI Overviews appear at the top of approximately 42% of UK Google searches, above the traditional blue link results.
- Being cited in an AI Overview drives brand visibility even when users don't click - and sometimes does drive clicks to your site.
- The pages most commonly cited are those that already rank in the top 10-20 for a query - so traditional SEO remains the foundation.
- The additional factors that determine which top-ranking pages get cited: E-E-A-T signals, structured Q&A content, FAQ schema, and content freshness.
- Brands that invest in AI Overview optimisation now are establishing a competitive position that will be increasingly difficult to dislodge.
If you've searched Google recently, you'll have noticed something different at the top of the page. Before the traditional results, before the ads, there's often an AI-generated answer - a synthesised response drawing on multiple sources, with small numbered citations linking to the pages it drew from.
That's Google AI Overviews - formerly known as Search Generative Experience (SGE) - and it now appears in an estimated 42% of all UK Google searches, according to data from BrightEdge's 2025 Search Landscape Report.
For most websites, this represents both a threat and an opportunity. The threat: AI Overviews reduce click-through rates to the organic results below them. The opportunity: being featured as a cited source in an AI Overview drives brand visibility at the very top of the page, ahead of even the number-one ranked result.
This guide explains exactly how to maximise your chances of being featured.
Google AI Overviews are generated by Gemini, Google's large language model, which synthesises content from across the web to compose a direct answer to a user's query. The key points:
AI Overviews draw from indexed, ranking pages Google isn't sourcing AI Overview content from the entire web equally. Research from Semrush and Ahrefs consistently shows that the pages cited in AI Overviews overwhelmingly come from the top 10-20 organic results for that query. If your page doesn't rank for the query, it almost certainly won't be cited.
They appear for informational queries, not transactional ones AI Overviews appear primarily for informational and research queries ("how does X work," "what is the best Y," "why does Z happen"). They appear less frequently for pure transactional queries ("buy X online," "X near me"), though this balance is shifting.
The model selects which pages to cite Even among the top-ranking pages, not all are cited. Google's model makes a selection based on factors including content structure, E-E-A-T signals, and the relevance of specific passages to the synthesised answer.
Users can expand citations Users can click to see more citations or click directly on a source. Click-through rates from AI Overview citations are lower than from traditional organic results, but they're not zero - and for many informational queries, appearing as a cited source may be more valuable for brand awareness than the click itself.
The honest answer: for informational queries, AI Overviews reduce clicks to traditional organic results. A Semrush study from late 2025 found that queries triggering AI Overviews see an average click-through rate reduction of 34% for organic results below the AI box.
However, the picture is more nuanced than pure traffic loss:
The strategic response isn't to abandon informational SEO. It's to optimise for AI Overview citation while continuing to build organic rankings - so you're present in both the AI-generated answer and the traditional results.
Based on analysis by SEO research firms including Semrush, Moz, and Search Engine Land, the following factors most strongly correlate with citation in Google AI Overviews:
As noted above, pages not ranking in the top 10-20 for a query are almost never cited. This means the foundation of AI Overview optimisation is standard SEO: technical health, quality content, and authoritative backlinks.
Pages ranking in positions 1-3 are cited most frequently. But pages in positions 4-20 are cited too - especially when they contain particularly clear or authoritative answers to specific sub-questions within the query.
Experience, Expertise, Authoritativeness, and Trustworthiness signals have a strong correlation with AI Overview citation rates. In practice this means:
Google's quality rater guidelines - which inform how its AI evaluates content - place significant weight on these signals.
The Gemini model is looking for passages it can extract and synthesise. Content structured as clear questions and direct answers is much easier to extract from than dense prose that requires interpretation.
Concretely: use your H2 and H3 headings as questions. Open the answer in the first sentence below the heading. Keep opening answers to 40-80 words before expanding into detail.
"The pages that get cited in Google's AI Overviews the most consistently aren't always the most comprehensive - they're the ones with the clearest, most directly answerable sentences. Google's model is pattern-matching for quotable passages."
- Tom Capper, Principal Search Scientist at Moz
Pages with FAQ schema markup are cited in AI Overviews at a meaningfully higher rate than those without. FAQ schema explicitly signals to Google's crawlers that the page contains question-and-answer content - which maps directly to how AI Overviews are constructed.
Adding FAQ schema to any page with a question-and-answer section is one of the highest-impact, lowest-effort AI Overview optimisations available.
Google AI Overviews show a notable preference for recently published or recently updated content. Pages that haven't been updated in 18+ months are cited less frequently, even when they rank well.
Regular content refreshes - adding new statistics, updating examples, adding a new section to reflect current developments - signal recency without requiring a complete rewrite.
Long paragraphs of unbroken text are harder for AI models to extract from. The formatting patterns most associated with AI Overview citation:
Here's how to approach AI Overview optimisation systematically across your site.
Start with queries where you already rank in positions 1-20 but aren't currently being cited in AI Overviews. Use Google Search Console to find your top-ranking informational queries. Then manually search those queries (in an incognito window) to see which ones trigger AI Overviews and whether you're cited.
Prioritise queries where:
These are your highest-leverage optimisation targets.
For each target page, assess:
Make a prioritised list of changes needed for each page.
For most pages, the changes are structural rather than substantive:
Across your site:
Check your target queries weekly for AI Overview appearances and citations. This is currently a manual process for most businesses - the tooling for AI Overview citation tracking is still maturing, though platforms like Semrush and Ahrefs are adding these features incrementally.
Track: citation rate for target queries, organic traffic from AI-Overview-triggered queries (via Search Console), and branded search volume (a proxy for brand visibility from non-click AI Overview appearances).
Google AI Overviews is one of several AI search channels worth optimising for. Here's how it compares:
| Platform | Traffic Volume | Citation Style | Key Optimisation Factors |
|---|---|---|---|
| Google AI Overviews | Highest | Inline citations with links | Organic ranking + E-E-A-T + structure |
| Perplexity AI | Growing | Full source list | Statistics, quotes, recency |
| ChatGPT web browsing | Growing | Paraphrased with links | Authority, clarity, comprehensiveness |
| Bing Copilot | Moderate | Inline citations | Bing ranking + structure |
For a complete guide to the broader landscape of AI search optimisation, see our GEO SEO guide.
Trying to game the system with thin "answer content" Some sites have attempted to create large volumes of thin, question-and-answer pages specifically targeting AI Overview appearances. This consistently backfires - Google's quality systems identify and devalue thin content quickly, and the pages rarely rank well enough to be considered for citation.
Ignoring technical SEO AI Overview optimisation is irrelevant if your pages can't crawl, index, and rank. Broken canonical tags, slow page speed, and crawl errors all prevent you from reaching the top 20 organic positions that are the prerequisite for citation.
Optimising only for AI Overviews and ignoring organic results AI Overviews exist above organic results - but organic results still drive the majority of search traffic for most sites. Pursue both in parallel.
Can I opt out of having my content appear in Google AI Overviews?
Not fully. You can use the nosnippet meta tag to prevent your content from being used in featured snippets and some AI-generated responses, but this also removes you from traditional featured snippets. It's a nuclear option that most businesses should avoid.
Does being cited in an AI Overview drive traffic?
Sometimes. Users can click the numbered citations in AI Overviews to visit the source. CTR from these citations is generally lower than from traditional organic results, but pages appearing as the primary citation (citation 1 or 2) do receive measurable click traffic.
Does AI Overview optimisation work for local businesses?
Local and service-area queries trigger AI Overviews less frequently than informational queries. Local businesses are better served by optimising their Google Business Profile, earning local reviews, and ensuring NAP (name, address, phone) consistency across directories.
How quickly does AI Overview optimisation take effect?
Schema changes and structural improvements can be picked up by Google's crawlers within days to weeks. Appearing in AI Overviews may take longer - it depends on when Google recrawls your pages and updates its model's understanding of your content.
Will Google AI Overviews continue to expand?
All available evidence suggests yes. Google has been expanding AI Overview coverage since the SGE rollout began in 2023 and there's no indication of a reversal. Businesses that build AI Overview visibility now are establishing an advantage that will compound as coverage grows.
Optimising dozens of pages for Google AI Overviews - rewriting headings, adding schema, strengthening E-E-A-T, keeping content fresh - is a significant undertaking. Most teams can't do it at the pace the opportunity demands.
Athenic is an AI-powered business platform that can manage the full AI Overview optimisation workflow for your site: auditing existing content, producing structured rewrites, implementing schema, and maintaining the freshness signals that keep you in Google's consideration set.