How to Get Cited by Google AI Overviews

Nearly half of all Google searches now show an AI Overview at the top of the page. That is not a prediction. It is happening right now, according to Search Engine Journal. And those AI-generated summaries are cutting organic clicks by 34.5%, based on Ahrefs research.

If your content is not showing up in those answers, someone else's is.

I spent the last year running tests across multiple sites and industries. What gets cited, what does not, and why. The patterns are consistent. It is not random. And it is not just about ranking well.

Google's AI picks sources based on content quality, structured data, trust signals, and whether your page adds something new. Most articles about this topic tell you to "do good SEO" and move on. That is like telling someone to "just be good at business." It is true, but it does not help.

Here is what actually works. I have tested these steps across different industries, different site sizes, and different levels of competition. The framework holds up.

What are Google AI Overviews?

How do you get cited by Google AI Overviews? AI Overviews are AI-generated summaries at the top of Google Search results. They are powered by Gemini. They pull information from multiple web sources and combine it into a single answer with links to the original pages. Getting cited means Google's AI decided your content was worth referencing. That decision is based on quality, structured data, trust signals, and whether your page offers something the other pages do not. The goal is not just to rank. It is to become a source that Google's AI trusts enough to cite in its answer.

Google calls this process "grounding." The AI retrieves fresh information from the web, synthesises it, and attaches citations to the source pages. It is not pulling from a static database. It is actively searching and evaluating in real time.

Here is what the data says:

• AI Overviews appear for 9.46% of all keywords and 16% in the U.S. on desktop (Ahrefs)
• They show for 54.61% or more of all Google searches by volume (Ahrefs)
• 7 out of 10 people never read past the first third of an AI Overview
• AI Overviews have grown 116% since the March Core Update (Ahrefs)
• They trigger more for informational queries, longer search queries, and queries with higher search volumes

I wrote a separate piece on Google AI Mode if you want to understand how Google's AI search features work in more detail.

How do AI Overviews select sources?

This is the part most people get wrong. They assume AI Overviews just pick the top-ranking page. They do not.

The process works like this:

• You type a query into Google
• Google's language models figure out what you are actually asking
• Google fetches relevant content from its index, looking at related searches and what people tend to click on
• Gemini takes all that information and writes a summary
• Google attaches citations to the pages it used
• The system learns from what people click (or do not click) and adjusts over time

The factors that determine which sources get cited:

• Content quality and relevance. Is the page accurate, thorough, and directly useful for the query?
• E-E-A-T signals. Does the content show real experience, expertise, authority, and trust?
• Structured data. Does the page have schema markup that helps Google understand what it is about?
• Information gain. Does the page say something new, or does it repeat what every other page says?
• Topical authority. Does the site have deep, consistent coverage of this topic?
• Page authority. Does the page have backlinks and external signals of trust?

Rich Sanger looked at Google's patent and found something interesting: "The AI Overview does not only provide documents from the top sources for that query. It seeks out diversity. If the top-ranked content for that query is homogenous, it will move on to closely related queries."

That is a important detail. Google's AI is not just copying the number one result. It is looking for the most useful, diverse information it can find. If the top 10 results all say the same thing, the AI will look elsewhere.

I wrote more about how search is shifting in my article on how SEO changes when search engines become answer engines.

The framework: 7 steps to get cited by AI Overviews

This is the framework I use with my clients. It is not theory. It is based on what I have seen work across multiple sites.

Step 1: Rank on page 1 first

According to Ahrefs, 76% of AI Overview citations come from top-10 ranking pages. Traditional SEO is still the foundation. You cannot get cited if Google does not know you exist.

Before you worry about AI optimisation, check:

• Is the page indexed?
• Does it rank on page 1 for the target keyword?
• Is the technical foundation solid (fast, mobile-friendly, crawlable)?
• Does it have relevant backlinks?

If the answer to any of these is no, fix that first. AI Overview optimisation does not work on top of a weak foundation.

This is the step people want to skip. Everyone wants the AI shortcut. There is not one. If your page does not rank well in traditional search, it is not going to get cited by AI Overviews.

Step 2: Implement structured data

Structured data helps Google's AI understand what your page is about. The schema types that matter most for AI Overview citations:

• FAQPage schema. Maps directly to question-based queries, which trigger AI Overviews more than any other query type
• Article schema. Defines who wrote it, who published it, and when. Feeds directly into trust signals
• Organization schema. Tells Google who your brand is and connects it to your other properties
• BreadcrumbList schema. Shows how your content is organised

I wrote a detailed guide on schema markup for AI search if you want the implementation details.

Step 3: Provide information gain

This is the step most people miss. And it is the one that matters most.

Google's AI looks for content that adds something new. If your page says the same thing as the other 10 results, the AI has no reason to pick you. It already has that information.

What counts as information gain:

• Original research. Data you collected. Surveys you ran. Experiments you conducted.
• First-hand experience. Real examples from your work. Case studies with actual numbers.
• Unique perspectives. Insights that come from doing the work, not from reading about it.
• Specific statistics with sources. Numbers that add credibility.

Google's patent confirms this. The AI actively looks for diversity. If the top results are all saying the same thing, it moves on to find something more useful.

This is why case studies and original data perform so well. They provide something that does not exist on every other page. If you have real numbers from your own work, publish them. If you have insights from years of experience, write them down. That is what makes your content worth citing.

I see this play out constantly. Two pages cover the same topic. One has generic advice anyone could write. The other has specific numbers from real campaigns, screenshots of actual results, and insights that only come from doing the work. The second page gets cited. Every time.

Step 4: Structure content for extraction

AI Overviews pull information out of your page. If your page is hard to pull from, it will not get cited.

Make it easy:

• Clear headings that match how people ask questions
• Concise definitions in the first 40-60 words of each section
• Comparison tables that AI can parse
• FAQ sections with self-contained answers
• Bullet lists for processes and steps

The format matters. A wall of text is hard for AI to extract. A page with clear headings, short definitions, and organised lists is easy. Think about how a machine would read your page, not just how a human would.

Step 5: Build topical authority

AI Overviews prefer sources from sites that know what they are talking about. That means depth, not breadth.

• Create content clusters. One pillar page covering the main topic, supported by detailed articles on subtopics
• Internal linking. Connect related content so Google sees the relationships
• Consistent entity signals. Your brand, your authors, your organisation should be described the same way everywhere
• Cover the full question space. Answer all the questions your audience asks, not just the main keyword

Topical authority is not about publishing more articles. It is about being the most thorough, most trusted source on a specific topic.

I worked with a client who had 200 blog posts covering their industry. None of them ranked well. We consolidated them into 40 deep, interconnected articles. Within three months, their AI Overview citations tripled. Depth beats breadth. Every time.

Step 6: Optimise for question-based queries

AI Overviews trigger more for questions than for simple keywords. People ask Google things like "how do I..." and "what is the best..." Those are the queries where AI Overviews show up most.

• Use question-format headings. "How do you...", "What is...", "Why does..."
• Answer directly in the first paragraph of each section
• Add FAQ sections with real questions and self-contained answers
• Think about how people actually ask. Not just the keyword, but the full question

I wrote more about this in my answer engine optimisation guide.

Step 7: Monitor and iterate

You cannot improve what you do not track. Check which of your pages get cited in AI Overviews:

• Manual testing. Search for your target queries. See if your content appears.
• Third-party tools. Ahrefs, Semrush, and similar tools can show which keywords trigger AI Overviews with your content
• Track patterns. Which formats, which topics, which structures get cited most
• Iterate. Use what you learn to improve other pages

For a proper framework on tracking AI visibility, see my article on how to measure AI citations.

Content formats that get cited most often

Based on what I have tested and observed:

FAQ sections are the most direct path to AI citation. They map to how people ask questions. Self-contained answers are easiest for AI to extract. FAQPage schema makes this even more effective.

Comparison tables give AI structured data it can pull directly. Side-by-side comparisons with specific numbers are highly citeable.

Step-by-step guides with numbered steps are easy for AI to extract. Clear, sequential instructions with HowTo schema are natural candidates.

Definition paragraphs that explain a concept in 40-60 words at the start of a section get cited frequently. The definition should stand on its own.

Data-rich content with statistics, cited sources, and specific numbers gives AI something concrete to reference. Vague claims do not get cited. Specific numbers do.

Common mistakes that prevent AI Overview citations

Thin content

AI Overviews favour detailed, complete content. Pages with 300-500 words rarely get cited. For pillar topics, aim for 2,000+ words. Cover the topic properly, not superficially. If someone searches for a complete guide and finds a 400-word overview, they leave. And so does the AI citation.

No structured data

Without schema markup, your content is harder for AI to parse. Implement FAQPage, Article, and Organization schema at minimum. It takes an hour and makes a measurable difference.

Generic content without information gain

If your page says the same thing as every other page, there is no reason for AI to cite it. Add original data, real examples, specific numbers. Give the AI something it cannot find elsewhere.

Poor content structure

Walls of text are hard to extract. Use headings, bullet lists, tables. Structure your content around questions and answers. Make it easy for a machine to pull out the key information.

Ignoring trust signals

Author credentials matter. Source citations matter. Transparent authorship matters. Google's AI evaluates whether your content is trustworthy before it cites you. If the trust signals are missing, the citation goes to someone else.

How to track your AI Overview citations

Google Search Console does not separate AI Overviews from organic search. All the impressions and clicks are lumped together. You need other tools.

Methods:

• Ahrefs Site Explorer. Enter your domain, go to Organic keywords, filter for AI Overview appearances
• Ahrefs Brand Radar. Track brand mentions in AI Overviews. Compare against competitors
• Semrush. AI Overviews tracking in the Positions report
• Manual testing. Search for your target queries. Check if your content shows up.

I wrote a complete guide on how to measure AI citations if you want the full tracking framework.

Industry-specific AI Overview patterns

AI Overviews do not trigger equally across industries. Understanding the patterns helps you prioritise.

Healthcare has the highest trigger rate at 88% of queries. That makes it both the biggest opportunity and the highest risk. Trust requirements are the strictest here.

Financial services triggers heavily with FCA-regulated content. FinancialProduct schema and clear author credentials are important.

Legal has a moderate trigger rate with SRA-regulated content. LegalService schema and proper disclaimers matter.

iGaming is growing fast with UKGC compliance requirements. Responsible gambling messaging and geo-targeted content are non-negotiable.

I cover all of this in my SEO for regulated industries guide.

AI Overviews do not replace SEO. They add to it.

AI Overviews are not the end of traditional SEO. They are a new layer on top of it. The fundamentals still matter. Crawling, indexing, technical health, content quality, backlinks. All of it still counts.

But the definition of success is changing.

Old model: rank and get clicked.

New model: rank, get cited, and get recommended.

The companies that win are the ones that optimise for both. They build technically sound sites. They create genuinely useful content. They implement structured data. And they make their content easy for AI to extract and reference.

This is not a threat to people who have been doing SEO well. It is an opportunity. If you have been creating high-quality, well-structured content, you are already in a good position. The framework here just helps you make the most of it.

The companies that figure this out early will have a real advantage. While their competitors are still trying to understand why traffic is dropping, they will be showing up in the AI-generated answers that more and more buyers are reading before they ever click a search result.

The shift from traditional search to AI-powered answers is not something to fear. It is something to prepare for. The framework here is the preparation.

For help with AI Overviews, structured data, and AI visibility, my LLMO services cover all of it.