How to increase visibility and get recommended in AI search engines (GEO)
How do you improve your visibility in AI search engines like ChatGPT, Google’s AI Mode, Copilot, and Perplexity?
Hi there. I’m Eric Waterschoot. And as an AI strategist, I get this question almost daily.
While Google is still the dominant way people discover information, a major shift is underway. Large Language Models (LLMs) like ChatGPT, Google’s AI Mode, Copilot, Perplexity, and DeepSeek are increasingly used as primary discovery engines, not just assistants.
Users now expect:
- Direct answers instead of lists of links
- Contextual explanations instead of keyword matches
- Trusted recommendations instead of endless scrolling
For website owners, this means one thing: SEO is evolving into GEO (Generative Engine Optimization).
But GEO does not replace SEO. It builds on it.
From SEO to GEO: What actually changed
GEO is often described as “the next version of SEO,” but that framing is misleading.
SEO is not going away.
Generative engines still depend heavily on the same foundations that power traditional search.
The difference is how content is selected and presented.
Traditional SEO aims to:
- Rank pages in search results
- Capture clicks through visibility
GEO aims to:
- Get cited or referenced directly in AI answers
- Become a trusted source when AI systems summarize a topic
In practice, GEO builds on SEO, adding an extra layer focused on authority, clarity, and citation-worthiness.
How do AI search engines crawl and index content?
AI search tools don’t necessarily use a “traditional crawl + index + rank” process exactly like Google. Instead, they combine selective crawling, curated indices, and real-time retrieval to generate answers.
Unlike Google that relies on Googlebot to continuously crawl and index vast portions of the open web, most AI-powered search tools maintain smaller, higher-quality indices.
In addition to stored indices, many AI systems also use on-demand retrieval. When a user asks a question, the AI may fetch fresh content in real time, evaluate multiple sources, and synthesize an answer (sometimes even if the content isn’t permanently indexed).
Different platforms approach this hybrid model in distinct ways:
- Perplexity AI uses PerplexityBot to crawl selected sources and also performs live retrieval at query time. Its crawling behavior can sometimes bypass robots.txt and other traditional restrictions.
- Microsoft Copilot relies on Bing’s standard index. This means content visibility in Copilot is largely dependent on how well a site is indexed and ranked in Bing.
- Google’s AI Overviews and AI Mode build directly on Google’s traditional index combined with live AI insights.
- ChatGPT uses OpenAI crawlers (GPTBot for training-related crawling and OAI-SearchBot for search indexing), both of which respect robots.txt. Blocking these bots can limit indexing and citation in ChatGPT Search (although content may still appear in responses if fetched live at a user’s request).
So, to ensure your content is discoverable in AI search:
- Allow AI crawlers access – Check your robots.txt and ensure essential content isn’t blocked.
- Enable key bots – Permit crawlers like PerplexityBot, GPTBot, OAI‑SearchBot, Googlebot, and Bingbot to index your content.
Ensure proper page indexability – Confirm canonical tags, remove unnecessary 404s or noindex directives, and maintain clean site structure.
Use SE Ranking’s Website Audit tool to detect technical SEO issues that might prevent AI search engines from properly indexing your content.
However, crawlability alone does not lead to AI visibility.
SE Ranking’s research shows that files like LLMs.txt currently have little to no influence on how tools like ChatGPT or Google’s AI Mode select and cite content. So, technical access is a prerequisite, but it’s not, by itself, a lever for AI visibility.
How to check your visibility in AI search engines
AI tools like ChatGPT, Copilot, and Perplexity AI generate answers based on databases, live search results, memory, and past interactions. This means your site’s visibility can be influenced in ways you might not expect. Always test anonymously and online to get a more realistic view:
- Use an incognito window instead of logging into ChatGPT, Copilot, or Perplexity AI.
- For DeepSeek, create a new account with no search history.
But manual testing has limits. AI tools may still personalize answers based on session context, model updates, and prompt structure, so results still vary.
That’s why SE Ranking developed tools like the ChatGPT Visibility Tracker, AI Overviews Tracker, AI Mode Tracker, which can be highly valuable for AI and SEO strategists like me. Specifically, they allow you to:
- Monitor which queries trigger your brand
- Track citations over time
- Compare your visibility against competitors
- Identify your most cited pages
This moves GEO from guesswork to measurable optimization.
What actually drives AI citations (based on data)
To answer this question, SE Ranking recently analyzed:
- 129,000 domains & 216,000 pages for ChatGPT
- 295,000 domains & 2.3M pages for AI Mode
- Hundreds of thousands of AI prompts
- Over 50 ranking factors using XGBoost and SHAP analysis
And here are the factors that have the most impact on citations in ChatGPT and AI Mode answers.
1. Domain authority
Authority is the strongest common denominator across AI platforms.
Sites that:
- attract quality backlinks
- earn references from trusted domains
- demonstrate long-term credibility
are significantly more likely to be cited by AI systems.
For ChatGPT in particular, referring domains are the #1 ranking factor.
To provide the example, sites with over 32K referring domains are 3.5x more likely to be cited by ChatGPT than those with up to 200 referring domains.
2. Google visibility
AI systems don’t operate in isolation from search engines.
Pages that already perform well in Google:
- rank higher organically
- attract consistent traffic
- maintain stable visibility
are far more likely to appear in Google AI Mode responses.
Research data shows that websites with over 1.16M visitors are 3x more likely to be chosen as a source by AI Mode than those with under 2.7K visitors.
This makes organic Google traffic the #1 predictor of AI Mode citations.
3. Brand strength and recognition
AI systems favor recognizable brands.
Strong signals include:
- Homepage traffic
- Brand search volume
- Brand mentions on Quora and Reddit
- Presence on review platforms (Trustpilot, G2, Capterra, Yelp)
For instance, domains with millions of brand mentions on Quora and Reddit have roughly 4x higher chances of being cited by ChatGPT than those with minimal activity. For AI Mode, reaching the point of 240K brand searches increases your chances of getting cited by around 33%.
4. In-depth, structured, regularly updated content
Pages that consistently perform well:
- cover a topic end-to-end
- address related sub-questions
- review and update important pages regularly
- use H2s for main themes, H3s for specifics
- use lists and tables where appropriate
- incorporate expert quotes, statistical insights, and so on
generally have higher chances to get cited in AI answers.
The research shows that long-form content with over 2,900 words performs best for ChatGPT. For AI Mode, this starts from 2.300 words.
Plus, for both AI systems, updating content every 2-3 months is considered an important ranking factor.
5. FAQ sections and question-styles headings
Including FAQs inside the content helps AI identify clear answers.
- FAQ sections increase citation likelihood
- Question-style titles and H1s help, especially for smaller sites
Important:
FAQ schema markup itself has no measurable impact on AI citations. Structure beats metadata.
A practical GEO action plan
1. Build and reinforce authority signals
- Earn mentions and backlinks from reputable, relevant websites
- Clearly identify content authors and their expertise
- Support claims with credible sources, data, or references
2. Maintain organic visibility on Google
- Prioritize high-traffic pages that already attract significant organic traffic
- Improve average ranking positions of key URLs
- Maintain strong homepage visibility and brand signals
3. Create in-depth, well-structured content
- Aim for at least 1,500 words for most topics
- Use H1, H2, and H3 headings that AI can interpret
- Break your content into sections of 120-150 words between headings
4. Strengthen your brand presence beyond your website
- Participate in relevant discussions on platforms like Reddit and Quora
- Establish presence on review platforms like Trustpilot, G2, Capterra, Sitejabber, and Yelp.
- Share your content updates across all social platforms
- Engage with your community to show an active brand presence
Tools like Planable help maintain this consistency without overwhelming your team
What doesn’t work for increasing AI search visibility:
❌ LLMs.txt files
❌ Keyword-stuffed titles and URLs
❌ Relying on schema markup alone
❌ Thin or overly generic content
❌ Focusing only on one type of content
❌ Relying on a “trusted” domain zone like .gov or .edu
❌ Ignoring internal linking
❌ Overloading pages with ads or pop-ups
Bonus tip: Use AI follow-up questions to get insights
AI tools like Perplexity and Copilot generate follow-up questions based on search queries. This reveals content gaps that you can use as inspiration for new content:
- What related topics are users searching for?
- What FAQs are missing from my website?
- What details does AI point out that I haven’t covered?
- What would be a logical next step for my visitors?
SE Ranking’s Content Editor helps you optimize this content for traditional search so you can drive more organic traffic from Google.
