The idea for this research started from a claim by Anna York LI post.
She mentioned that brands with strong authority are the ones appearing more often in AI-generated search results.
At first, it sounds logical.
After all, search engines have always rewarded authority. So it feels natural to assume that AI systems would do the same.
But here’s the real question:
Does high authority actually make it easier to get cited in AI results? Or is something else driving those citations?
We decided to test this with real data.
Key Takeaways
Before we go deep, here are the main insights from our research:
- High authority is not required for AI citations
- Around 30% of cited pages had low authority (DA < 25)
- Pages that match user intent and semantic meaning perform better
- Structural clarity plays a major role in citations
This Uprankly guide explains how AI citations work and what factors influence them. Also, you will understand why authority alone is not enough.
How We Approached The Research
To avoid assumptions, we followed a structured approach.Also, we used SEMrush to get the SEO metrics and other data.
The goal was simple: Understand what actually gets cited inside AI-generated answers.
Step 1: Selecting the Queries
We started by selecting 25 queries.
All of them were related to:
- AEO (Answer Engine Optimization)
- GEO (Generative Engine Optimization)
- SEO
These are the areas where AI-generated answers are becoming more common.
Some of the queries we used:
- how to get cited in AI Overviews
- what are GEO optimization tips
- what is AEO in SEO
- how does Google AI Overview choose sources
- how FAQs help with AEO
etc.
We made sure the queries reflect real user intent and common search behavior.
Step 2: Analyzing AI Overviews
Next, we searched each query on Google and checked:
- Does Google show an AI Overview?
- What answer is being generated?
Whenever an AI Overview appeared, we carefully reviewed the response.
Then we collected all the pages that Google cited as sources.
In most cases, each answer included 3 to 5 cited pages.
These pages became the core of our analysis.
Step 3: Evaluating the Cited Pages
We didn’t stop at listing the sources.
We manually visited every cited page and analyzed it.
Our main question was simple:
Does this page actually support the AI-generated answer?
Based on this, we created three categories:
- Grounded: The page clearly supports the answer
- Partial: The page supports some parts of the answer
- Non-grounded: The page does not support the answer
This helped us understand how accurate and relevant the citations really are.
Does Domain Authority Really Matter for AI Citations?
No, domain authority does not strongly determine AI citations. AI prioritizes relevance, structured content, and intent match.
This was the most important part of our research.
Because the whole idea started with authority.
So we analyzed the authority level of each cited domain.
And the result was interesting.
- Around 30% of the cited websites had a domain authority below 25
This means:
- Many cited sites were not highly authoritative
- Their backlink profiles were relatively weak
Yet, they still appeared in AI-generated answers.
So, it changes a common assumption.
Authority is helpful, but it is not the main factor.
AI systems are not just picking the biggest brands.
Instead, they are picking content that:
- Answers the query clearly
- Matches user intent
- Provides useful and structured information
So, What Actually Drives AI Citations?
AI citations are driven by clear answers, strong structure, deep content, semantic relevance, and alignment with user intent rather than authority alone.
Based on everything we observed, here are the real drivers:
1. Clear and Direct Answers: Content should solve the query without confusion.
2. Depth of Information: Surface-level content is not enough.
AI prefers pages that explain things properly.
3. Strong Structure: Headings, lists, and FAQs make content easier to extract.
4. Intent Alignment: Content must match what the user is trying to achieve.
5. Semantic Coverage: Using related concepts helps AI understand the topic better.
Are AI Citations Based on Exact Keyword Matching?
No, AI citations are not based on exact keyword matching. Citations rely on semantic relevance, context understanding, and how well content answers the query.
One of the most surprising findings came from here.
We expected AI to rely heavily on exact phrases.
But that didn’t happen.
In most cases:
- The exact wording from the AI answer was not found in the cited pages
- Instead, the pages covered the topic in detail
- AI summarized the content into a concise answer
This means:
- AI does not look for exact matches
- AI looks for meaning and context
So even if your content does not match the query word-for-word, it can still get cited.
As long as it answers the question clearly and deeply.
What Kind of Content Gets Cited?
Well-structured content with clear headings, concise answers, bullet points, FAQs, and strong relevance to user intent is more likely to get cited.
We noticed a strong pattern across almost all cited pages.
They were not random blog posts.
They were well-structured and easy to scan.
Here’s what most of them had in common:
- Question-based titles
- Clear and logical subheadings
- Bullet points for quick reading
- Tables for organized information
- FAQ sections answering related questions
These elements make content easier to understand.
But more importantly, they make it easier for AI systems to:
- Identify key points
- Extract relevant answers
- Build summaries
In short: Well-structured content increases your chances of being cited.
The Role of User Intent and Semantic Relevance
Another key insight was how strongly AI focuses on intent.
The cited pages did not just mention the topic.
They fully addressed what the user was looking for.
For example:
- If the query was “how to get cited,” the page explained steps
- If the query was “what is AEO,” the page gave a clear definition
Beyond that, the pages also covered related ideas.
This is where semantic relevance comes in.
Instead of repeating the same keyword, they included:
- Related terms
- Supporting concepts
- Deeper explanations
This helps AI understand the full context.
Final Thoughts
We started with a simple assumption: High authority leads to easy AI citations
But the data tells a different story.
Authority still matters but it is not the deciding factor.
AI is shifting the focus.
From “who is the biggest brand” to “who provides the best answer”
That means even smaller sites have a real opportunity.
If your website content is in-depth matching user intent, semantically relevant, and well-structured for AI extraction, you can compete with high-authority domains and get cited.
This research highlights a larger trend.
Search is no longer just about ranking.
It is about being understood and selected by AI.
And that requires a different approach:
- Focus less on keywords alone
- Focus more on meaning and usefulness
Because in AI search:
The best answer wins, not just the strongest domain