How Google AI Overviews Work - And How to Get Featured in Them

Google AI Overviews now trigger on nearly half of all tracked search queries. AI Overviews trigger on nearly half of all tracked queries - a 58% increase year over year. If you rank first for a keyword that produces an AI Overview, you might collect less than half the clicks that same position earned you in 2024. As of December 2025, AI Overviews reduce the organic click-through rate for position one content by 58%.

That number isn't speculation. It comes from Ahrefs' analysis of 300,000 keywords comparing December 2023 data against December 2025 data. They compared the clickthrough rates for both samples for December 2023 (pre-AI Overviews) and December 2025. Seer Interactive corroborates the trend with its own study of 25.1 million organic impressions, showing that organic click-through rates for informational queries featuring Google AI Overviews fell 61% since mid-2024.

But here's the data point that should redirect your strategy, not paralyze it: cited brands receive approximately 35% more organic clicks and 91% more paid clicks compared to uncited competitors appearing on the same results page. The game has shifted from ranking to getting cited. This post breaks down exactly how AI Overviews select sources, what changed in 2025 and early 2026, and the specific steps that earn your content a spot inside the answer.

What AI Overviews Actually Are (and Aren't)

Google AI Overviews are AI-generated summaries that appear at the top of Google search results pages for certain queries. They were formally launched in May 2024, following an earlier experimental phase known as Search Generative Experience (SGE). Think of them as featured snippets fused with a research assistant. Rather than pulling a single passage from one page, Google's Gemini model reads content from multiple web pages, synthesizes the information, and produces a concise answer directly in the search results page.

The scale is no longer experimental. As of 2026, AI Overviews are available in over 200 countries and more than 40 languages. Google CEO Sundar Pichai confirmed 2 billion monthly users on the Q2 2025 earnings call. Google AI Mode - the fully conversational companion feature - crossed the 75M daily active user threshold by early 2026, representing a 4x increase since its May 2025 launch.

Not every query triggers an Overview. Informational queries trigger AI Overviews most frequently. Transactional and local queries trigger them less often. Branded searches almost never get an AI Overview. This matters for prioritization: if your keyword portfolio skews informational, the urgency is higher. If you sell products through e-commerce, only 4% of e-commerce searches trigger AI Overviews in 2026 - giving you more breathing room, though not immunity.

The RAG Pipeline: How Google Selects Sources for AI Overviews

Understanding the technical mechanism separates useful optimization from cargo-cult tactics. AI Overviews use a retrieval-augmented generation (RAG) process. Instead of relying solely on the model's internal training, which risks hallucinations, Google's systems pull in real content from the web.

The pipeline works in stages. Google determines if a query warrants an AI Overview based on complexity and intent. The system performs a traditional search to identify high-quality, relevant documents. The Gemini model extracts key facts, entities, and relationships from these documents. The model generates a coherent summary, ensuring that every claim is grounded in the retrieved sources. The system maps specific parts of the generated text back to the source URLs, creating the citation cards.

What makes this different from a featured snippet is the query fan-out step. Google has confirmed that its system performs a "Query fan-out" whenever a user searches and AI is triggered. This is when the initial query is split into multiple related sub-queries. The pages that appear most often within those sub-query SERPs then get cited in the AI Overview.

Consider a query like "best electric SUV." Google doesn't just evaluate the top 10 results for that phrase. It secretly also searches for "electric SUV safety ratings," "EV charging infrastructure," "Tesla Model Y vs competitors," and potentially a dozen more sub-queries. Your page gets cited not because it ranks for one term, but because it shows up across multiple fan-out queries. This is why the relationship between organic ranking and AI citation has weakened dramatically. Google is selecting far fewer pages straight from the original SERP - approximately 76% in July 2025 versus approximately 38% today. A separate BrightEdge analysis puts the overlap even lower, at approximately 17%. Ranking on page one is no longer a reliable proxy for earning an AI Overview citation.

Why the Rules Changed: From Ranking to Topical Coverage

The shift from 76% to 38% top-10 citation overlap happened in under a year. Three forces explain the speed. Gemini 3 and more aggressive fan-out. With the arrival of Gemini 3, fan-out queries may be playing a bigger role in source selection. It could be expanding queries more aggressively than before, pulling from related SERPs where fewer pages rank for the original query. Google hasn't publicly confirmed the change, but the timing aligns with Ahrefs' data shift. Passage-level extraction, not page-level. AI Overviews don't reference entire articles. AIO doesn't pull entire articles. It pulls specific passages of 134–167 words that answer a specific sub-question, providing a complete answer without requiring additional synthesis. Your page might have a brilliant section buried under three unrelated topics. If that section stands alone as a clear answer to a fan-out query, it can get cited - even if the page as a whole ranks #30. YouTube's rising dominance. Among the AIO citations that didn't rank in Google's top 100 for the same keyword, 18.2% were YouTube URLs. YouTube accounted for 5.6% of all AI Overview citations in the dataset. Ahrefs also reported that YouTube is the most-cited domain in AI Overviews overall and has grown 34% over the past six months. If your brand has no YouTube presence, you're conceding citation opportunities to competitors or to platform content. The practical implication is uncomfortable but clear: SEO strategies focused exclusively on ranking for a single keyword are less likely to secure AI Overview citations. Instead, content must address a wider array of related topics and user intents captured by fan-out queries.

What Google Officially Says (And Where Practitioners Disagree)

Google's public guidance is deliberately minimal. The best practices for SEO remain relevant for AI features in Google Search. There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary. Their documentation states that to be eligible to be shown as a supporting link in AI Overviews or AI Mode, a page must be indexed and eligible to be shown in Google Search with a snippet. There are no additional technical requirements.

Google also claims positive traffic outcomes. When people click from search results pages with AI Overviews, these clicks are higher quality, where users are more likely to spend more time on the site. This "quality over quantity" framing directly contradicts what independent studies measure. Where practitioners disagree is on whether "no special optimizations" tells the full story. Google says best practices make your content "eligible" for citation, but eligible doesn't mean selected. The brands winning citations consistently go beyond baseline eligibility. They optimize for extractability, topical breadth, and entity signals. A balanced reading: Google isn't lying - traditional SEO fundamentals are genuinely foundational. But treating them as sufficient ignores what the citation data shows. Double down on technical excellence, high-quality content, and building authority. SEO is the price of entry for being considered by the generative engines that are built upon it. Entry is not the same as winning.

The Practitioner's Playbook: 7 Steps to Earn AI Overview Citations

1. Audit Your Keyword Portfolio for AI Overview Exposure

Before optimizing anything, determine which of your keywords trigger AI Overviews and whether you're currently cited. You can use the SERP features filter to see exactly where AI Overviews are triggering for your keyword rankings. Head to Ahrefs' Site Explorer, navigate to the Organic keywords report, and select AI Overview from the SERP features filter. Semrush and Conductor offer similar tracking. This audit tells you where traffic erosion is most severe and where citation opportunities exist. Segment your keywords into three buckets:

• No AI Overview: Traditional SEO applies. Protect these rankings.
• AI Overview present, you're cited: Monitor and defend this position.
• AI Overview present, you're not cited: Highest optimization priority.

2. Build Content Around Fan-Out Query Clusters, Not Single Keywords

The fan-out mechanism means Google evaluates your content against sub-queries you may never have targeted. The more fan-out queries a URL ranked for, the more likely it was to show up in Google's AI Overviews. Surfer SEO's analysis of 173,000 URLs found that pages ranking for fan-out queries are 161% more likely to be cited. Practically, this means building topic clusters. Topic clusters are groups of interlinked webpages that work together to cover a core topic comprehensively. They're made up of a central pillar page and several cluster pages covering relevant subtopics. Topic clustering helps you to address multiple queries that may be generated through relevant query fan-outs.

Identify fan-out queries using People Also Ask data, Google Search Console long-tail queries, and tools like AlsoAsked or Surfer's AI Tracker. Gather real user questions from your sales calls and support tickets, as well as forums like Reddit and Quora. Map those questions to content gaps, then fill them.

3. Structure Every Section as a Standalone, Extractable Answer

Remember: the AI pulls passages, not pages. Content structured so that each section under a heading functions as a standalone answer to a discrete question aligns with this extraction pattern. The "answer-first" format - direct answer in 1–2 sentences, then explanation, then supporting details - mirrors how AIO constructs its own output.

Specific formatting guidance:

• Lead with the answer.

Open every page with a 40–80 word "Quick answer" that directly addresses the core query, then expand with context.

• Use question-based H2s that mirror how people actually search.

Structure your H2s as actual questions that mirror real user searches.

• Keep extractable passages self-contained. Each section under an H2 should make sense without the reader having seen the preceding section.
• Use lists, tables, and comparison formats where the content warrants them.

Direct-answer paragraphs of 40–60 words, definition sentences, numbered process lists, comparison tables, and FAQ sections are the formats most frequently cited in AI Mode.

4. Strengthen E-E-A-T Signals Across Your Entire Site

AI Overviews inherit Google's quality evaluation framework. E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) remains critical for GEO. Content with transparent author bios, reputable citations, and consistent updates often outranks shallow material.

For YMYL topics, Google draws from an even smaller pool. "In the healthcare category where accuracy and trustworthiness are paramount, Google is increasingly showing search results from just a handful of websites. Content from authoritative medical research centres account for 72% of AI Overview answers."

Actionable E-E-A-T steps:

• Attribute content to credentialed authors with visible bios and linked professional profiles.

Include examples of first-hand experiences, case studies, and original perspectives that commodity content cannot replicate. - Cite reputable external sources with live URLs. Footnote statistics with live URLs - LLMs reward transparent citations.

• Keep content updated. Stale pages with outdated data lose trust signals for both traditional rankings and AI citation.

5. Implement Strategic Schema Markup

Schema markup is not a magic switch for AI citations, but the evidence supports it as a meaningful signal. Google's Search team said that structured data gives an advantage in search results. Microsoft Bing confirmed that schema markup helps its LLMs understand content for Copilot.

A nuanced view: A December 2024 study found no correlation between schema markup coverage and citation rates. Sites with comprehensive schema didn't consistently outperform sites with minimal or no schema markup. However, attribute-rich schema earns a 61.7% citation rate, but generic, minimally populated schema actually underperforms having no schema at all. Implementation quality matters more than implementation quantity. Priority schema types for AI visibility:

• FAQPage -

FAQ schema visibility decreased in Google search results, but FAQ schema importance skyrocketed for AI search visibility.

• Article/BlogPosting - establishes content type, authorship, and publication date.
• Organization - anchors your entity identity in Google's Knowledge Graph.
• Person - connects authors to their credentials and external profiles.
• HowTo - maps step-by-step instructions in a format AI can directly extract.

Use JSON-LD format. Validate with Google's Rich Results Test. And critically: every schema type you add should reflect content that's actually on the page. Adding FAQPage schema to a page with no visible FAQ section violates Google's structured data guidelines.

6. Build Brand Entity Signals Beyond Your Own Website

AI Overviews don't evaluate your content in isolation. The more citations you have in that pool of data, the greater your chances of showing in AI Overviews. Alongside your link building strategy, build relevant brand mentions across the web: third-party websites, online forums, social media platforms, news publications.

YouTube deserves specific attention. Research of 75K brands revealed that mentions on YouTube - in video titles, transcripts, and descriptions - are the strongest correlating factor with AI Overview visibility. Creating video content that mentions your brand name, covers your expertise topics, and links back to your site builds citation signals that text-only strategies miss. Earn editorial mentions from industry publications. Participate in expert roundups. Contribute original data to forums where your audience asks questions. Each third-party mention corroborates the entity signals that Google's systems use to decide whether your content is citation-worthy.

7. Track AI Citation Performance as a Distinct KPI

Traditional rank tracking doesn't capture AI visibility. AI Mode and AI Overviews cite the same URL only 14% of the time. Even within Google's ecosystem, the two AI surfaces pull from different source pools. You need dedicated monitoring. Tools to consider:

• Ahrefs Brand Radar - tracks AI Overview citations and brand mentions across AI platforms.
• Semrush AI Visibility Toolkit - monitors share of voice across non-branded queries in AI responses.
• Seer Interactive's Generative AI Tracker - categorizes queries by AIO presence and citation status.
• ZipTie, Profound, or Surfer AI Tracker - purpose-built tools for cross-platform AI citation monitoring.

Your SEO strategy cannot rely on changing your targeting to avoid AIOs. You'll need a strategy that reframes your KPI as maintaining your share and visibility, rather than simply driving traffic. Measure citation share alongside traditional metrics. Report on it monthly.

What AI Mode Means for the Road Ahead

AI Overviews sit alongside traditional search results. AI Mode replaces them entirely. Google AI Mode is an end-to-end AI search experience powered by Gemini that provides comprehensive answers with citations but eliminates the traditional 10 blue links. In AI Mode, you either get cited - or you don't.

The two features are converging. Since January 2026, Google now lets users tap from an AI Overview straight into an AI Mode chat to go deeper. This creates a two-stage flow: AI Overviews as the entry point, AI Mode as the research space. Optimizing for AI Overviews today builds the foundation for AI Mode visibility tomorrow. Google AI Mode uses the same query fan-out mechanism but more aggressively - AI Mode uses a fan-out technique that issues multiple queries simultaneously (up to 16 searches). The content architecture required for one feeds directly into the other: topic clusters, extractable passages, strong entity signals, and structured data. The paradigm shift isn't subtle. If you're not visible in AI Mode, you're not visible. That may be overstated for 2026, given that AI Mode still requires users to actively choose a different tab. But the trajectory is clear. Building citation-worthy content now means you're positioned when - not if - conversational search becomes the default. --- The core argument hasn't changed since Google launched its first featured snippet a decade ago: be the most useful, most trustworthy answer to a question. What has changed is the mechanism through which "useful" is evaluated. AI Overviews decompose queries, extract passages, and synthesize answers from across the web. Your content must be structured for that extraction, authoritative enough to survive that evaluation, and comprehensive enough to surface across multiple fan-out queries. Organizations that treated AI Overviews as a passing experiment have already lost ground. The common thread across successful adaptations: organizations that shifted to proactive AI visibility strategies within 6 months of AI Overview expansion significantly outperformed peers. The window for proactive adaptation is still open, but narrowing. Start with your keyword audit, restructure your highest-value pages for extractability, and build the entity signals that make AI systems trust your brand enough to cite it by name.