Why AI Ignores You — And How to Get Cited by ChatGPT, Perplexity, and Google AI Overviews

I asked ChatGPT to name the top experts in a client’s field. It listed five people. My client wasn’t one of them.

She’d been in that field for 14 years. Two books. Keynote speaker at her industry’s biggest annual conference. More credentials than anyone ChatGPT named. But AI had never heard of her.

Not because she wasn’t good enough. Because she wasn’t visible in the way AI systems understand visibility.

Here’s the scale of what that means. ChatGPT now handles over 2 billion queries daily. Perplexity and Google AI Overviews handle hundreds of millions more. In the time it took you to read those last two sentences, tens of thousands of people asked an AI a question in your area of expertise. Some of those responses cited an expert by name. It almost certainly wasn’t you.

That’s fixable. This guide explains exactly why AI skips you, what signals actually trigger citations, how each major platform decides differently, and what to do in the next 90 days to move from invisible to cited by name.

Key Takeaways

• Entity density optimization delivers a 292% lift in AI citation probability. That’s not a typo. (iPullRank, 2026 study of 79,000+ URL-query pairs).
• Only 11% of domains are cited by both ChatGPT and Perplexity. Each platform has completely different citation logic. You need platform-specific strategy.
• 44.2% of all AI citations come from the first 30% of a document. Your introduction is your most valuable real estate.
• Position 1 CTR drops 58% when a Google AI Overview appears (Ahrefs, Dec 2025 study of 300,000 keywords). Being cited in AI answers is now worth more than ranking first.

Why AI Systems Cite Some Experts and Not Others

The citation selection process isn’t random. A 2026 iPullRank study of 79,000+ URL-query pairs found that mid-tail queries with optimized entity density see a 292% lift in AI citation probability (iPullRank, 2026). That number points to something most AI visibility guides completely miss: entity density, not content volume, is the primary citation lever.

Here’s the thing: AI platforms don’t cite “good content.” They cite sources they’ve learned to associate with credible, domain-specific expertise. ChatGPT, Perplexity, and Google AI Overviews each use different retrieval architectures, but they share the same underlying logic.

For thought leaders, that creates a two-layer problem.

The first layer is entity recognition: does the AI know who you are as a distinct, credible Person entity? The second layer is content extraction: can the AI pull accurate, citable answers from your published content?

My client with 14 years of credentials had neither layer fully built. She had content. Plenty of it. But her entity wasn’t recognized by AI systems as authoritative. And the content she did have wasn’t formatted for extraction.

Most thought leaders are in exactly the same position. They have expertise. They lack AI visibility. Missing either layer reduces your citation probability significantly. But here’s the good news: both layers are buildable, and the process is more systematic than you’d think.

The 5 Signals AI Systems Use to Select Citations

Across all three major platforms, five signals consistently determine whether your name gets cited or someone else’s does. Get all five working together and citation probability compounds. Miss any one of them and you’ve got a gap a competitor can fill.

Signal 1 — Entity recognition. Does the AI system have a model of who you are as a Person entity? Knowledge Graph presence, Wikidata entries, and consistent cross-source corroboration feed this layer. Professionals with Knowledge Panels are recognized entities. Those without are unresolved names. And unresolved names don’t get cited by name. They get cited as “experts say” or not at all.

Signal 2 — Topical authority. Is your name consistently linked to a specific, well-defined domain? “Jennifer Park, financial planner” who has published extensively on retirement planning for self-employed professionals will be cited on that specific topic. Generalists are harder for AI systems to cite confidently. Specificity is a citation advantage, not a limitation.

Signal 3 — Content extractability. Is your content formatted so an AI can pull a clean, self-contained answer? Answer-first paragraphs (40-60 words that directly answer the heading’s implied question) and FAQ schema are the two highest-impact formatting choices. The 44.2% first-30% rule is real. Your opening paragraphs are the most-cited part of every piece you publish. Write them that way.

Signal 4 — Source authority and freshness. AI systems weight recently updated content from recognized sources more heavily than older content from unknown sources. For Perplexity especially, publishing frequency is a significant citation signal. A site that publishes regularly gets cited more often than a static one with comparable authority. Turns out, consistency isn’t just advice for newsletters.

Signal 5 — Cross-source corroboration. Is the same information about you confirmed by multiple independent sources? AI systems learn to trust claims that appear consistently across multiple authoritative references. This is exactly why the entity corroboration strategy that builds your Knowledge Panel (15-30 independent sources) also builds your AI citation profile. The two problems have the same solution.

What Nobody Tells You About AI Citations: Each Platform Works Completely Differently

Here’s what nobody tells you about AI citations: each platform has completely different logic. Only 11% of domains are cited by both ChatGPT and Perplexity (Conductor, 2026). That means what earns you citations on one platform doesn’t automatically transfer. You need to understand the architecture of each one.

How ChatGPT Selects Citations

ChatGPT operates in two modes: from its training data (with periodic updates) and from live web browsing via Bing when citations are needed. For thought leaders, both matter. And they respond to different signals.

From training data, ChatGPT heavily favors Wikipedia and encyclopedic content. 47.9% of ChatGPT’s top citations come from Wikipedia-adjacent sources. If you have a well-populated Wikidata entry, or you’re mentioned in Wikipedia articles about your domain, ChatGPT’s training layer is far more likely to have your entity in its model.

From web browsing, ChatGPT uses Bing’s index. That means Bing SEO signals matter, not just Google. For your entity home and key content pages, make sure they’re indexed and ranking in Bing alongside Google. ChatGPT also favors “answer capsules”: content formatted as clear, direct responses to specific questions, with the answer stated in the first one to two sentences.

One more thing worth knowing: since June 2025, ChatGPT appends utm_source=chatgpt.com to citation links. Set up an “AI traffic” channel group in GA4 if you want to actually measure which content is being cited.

How Perplexity Selects Citations

Perplexity is a real-time retrieval system. It searches the web at query time rather than relying primarily on training data. That makes freshness a much stronger signal here than it is for ChatGPT.

The result? A site that publishes regularly gets cited more often than a static site, even if their domain authority is comparable. And Perplexity heavily cites Reddit — 46.7% of citations come from Reddit-type community content. For thought leaders, this means three specific actions matter:

• Direct forum participation: answer questions in Reddit communities, Quora, and industry-specific forums. Perplexity sees these as authentic, primary-source responses.
• Citation reciprocity: Perplexity prefers sources that cite other authoritative sources. Content that references peer-reviewed studies and established experts is treated as more credible than unsupported assertions.
• Freshness cadence: publish at minimum monthly, ideally bi-weekly, to maintain Perplexity’s freshness signals.

Which brings us to the third platform — and it works completely differently from both.

How Google AI Overviews Select Sources

Google AI Overviews pull from Google’s search index, but they apply a secondary selection layer beyond traditional ranking signals. These are the primary drivers of inclusion:

• Topical authority: Google’s March 2025 update explicitly stated that AI Overviews prioritize sources with demonstrated topical expertise and clear author credentials. Your About page, author bios, and Person schema need to clearly establish your topical domain.
• Content type preference: blog content, video content, and article content are the most-cited content types in AI Overview results (Conductor, 2026). Structured, long-form articles with clear H2 sections and FAQ schema outperform generic informational pages.
• Industry-specific density: healthcare content appears in AI Overviews in 48.75% of searches. Real estate appears at only 4.48%. Know your industry’s AI Overview density and set expectations accordingly.

How to Build Your AI Citation Foundation: The Entity Layer

Most AEO guides skip this part entirely. They jump straight to content formatting — answer-first paragraphs, FAQ schema, topic clusters — and miss the fact that AI systems are significantly more likely to cite content from recognized entities than from anonymous domain names. The entity layer isn’t optional. It’s the prerequisite.

Here’s what “entity layer” actually means in practice:

Our finding: The professionals who get cited most consistently in AI responses aren’t just prolific publishers. They’re recognized entities. When an AI system has both an entity signal (Knowledge Graph presence, Wikidata entry) AND extractable content from that entity, citation probability increases dramatically. Entity signals without content don’t generate citations. Content without entity signals gets cited without your name attached. You need both tracks running at the same time.

1. A Knowledge Panel. Your Knowledge Panel tells AI systems that Google has verified you as a recognized Person entity in your domain. Professionals with panels are cited by name. Those without are cited as “experts” or “sources” — no attribution, no brand building, no compound return on your content investment.

2. A Wikidata entity entry. Wikidata is weighted heavily by both ChatGPT (encyclopedic preference) and Google’s entity recognition systems. A Q-number entry with consistent attributes takes about an hour to create. It has outsized, disproportionate impact on citation recognition relative to the time it takes. Do this one first.

3. Twenty or more corroborating sources. The same source diversity that triggers a Knowledge Panel also teaches AI systems that multiple independent, authoritative voices recognize you as an expert. This cross-source consensus is how AI systems validate entity-to-domain associations. Think: podcast appearances, speaker profiles, association directories, editorial mentions.

4. Person schema on your entity home. Your website’s About page with full Person schema is the machine-readable declaration that connects your identity to your topical authority. The knowsAbout property is particularly important. It’s how AI systems map experts to topics during retrieval.

For the full entity-building process, see our guide to entity SEO for people.

How to Format Content So AI Actually Cites You

Once the entity layer is in place, content formatting is what determines whether your words get cited or your competitor’s do. Same entity signals, better-formatted content: you win. And the formatting rules are specific enough that you can apply them to every piece you’ve ever published.

The answer capsule format. Every H2 section should open with a 40-60 word paragraph that directly answers the heading’s implicit question. That paragraph needs to be self-contained — meaning an AI can lift it out of context and use it as a standalone response without losing meaning. Include one specific statistic with source attribution inside it.

Here’s the difference in practice:

Without answer capsule:

“Thought leaders have been trying to get cited by AI systems since the introduction of ChatGPT in late 2022. There are many different approaches and strategies…”

With answer capsule:

“Thought leaders get cited by AI systems through a combination of entity recognition and content extractability. A 2026 iPullRank study found that optimized entity density increases AI citation probability by 292% (iPullRank, 2026). The process requires two parallel tracks: building your Knowledge Graph entity record and formatting content for direct extraction.”

The second version is immediately citable. An AI can take those 52 words, attribute them to you, and drop them into a response. The first version gets skipped.

The 44.2% rule. 44.2% of all AI citations come from the first 30% of a document. Write your most citable, most expert content in your introduction and first major section. Don’t bury your key insights in section four, where almost no AI system will extract them.

FAQ schema. FAQ sections with FAQPage JSON-LD schema are one of the highest-impact AEO investments available. Every article you publish should have 3-5 FAQ items with JSON-LD markup. Google’s research indicates properly marked-up FAQ content is cited in AI Overviews significantly more often than unmarked content covering the same questions.

Topic cluster architecture. AI systems prefer to cite sources with comprehensive topical coverage. A professional who has published 10 related articles on executive coaching will be preferred over one with a single long post. Depth across a topic — not just within one article — signals domain authority to AI retrieval systems. And that’s not all: each article in the cluster increases the surface area for citation across different query types.

The 90-Day Plan to Go From Invisible to Cited

The fastest path to consistent AI citations combines entity foundation work with content optimization. Not one or the other. Both, in sequence. Here’s exactly what to do and when.

Month 1 — Entity Foundation

By the end of this month, AI systems will have something to recognize when your name appears. That’s the goal. Everything else builds on this.

• Audit your current AI citation profile: ask ChatGPT and Perplexity “Who is [Your Name]?” and “What experts should I follow for [your domain]?” Write down exactly where you stand.
• Create or optimize your Wikidata entity entry with consistent attributes.
• Implement Person schema on your entity home with knowsAbout mapped to your core topics.
• Update your entity home bio to include your qualifier stack consistently.
• Begin building toward 20 corroborating sources: start with LinkedIn, Crunchbase, 2-3 podcast appearances.

Our experience: Professionals who complete Month 1 before touching content see faster citation gains than those who optimize content first. The entity layer is the foundation. Without it, content optimization is building on sand.

Month 2 — Content Optimization

By the end of this month, your best existing content will be formatted for extraction. New content will be built citation-ready from the start.

• Audit your top 10 existing articles: do they open with answer capsules? Do they have FAQ schema? Retrofit the ones that don’t.
• Publish 2-3 new articles in your core topic cluster, each with answer-first H2 sections and FAQ markup.
• Begin participating in 2-3 Reddit communities in your area of expertise. Genuine expert responses, not promotion.
• Add disambiguatingDescription to your Person schema if you share a name with other professionals.

Month 3 — Authority Signals

By the end of this month, you’ll have measurable evidence of citation activity and a system to track it going forward.

• Pursue 2-3 Tier 2 editorial mentions: podcast appearances, guest articles in industry publications, speaker profiles.
• Validate that your entity appears in the Knowledge Graph API (search kgsearch.googleapis.com for your name).
• Set up GA4 AI traffic attribution with the utm_source=chatgpt.com channel group.
• Track citation frequency monthly: ask ChatGPT and Perplexity your domain’s key questions and note whether you’re cited. Every month, the answer should improve.

For the broader framework connecting your AI citation strategy to your Knowledge Graph presence, see the complete AEO guide for notable professionals. To understand how AI citation connects to your Knowledge Panel, see the guide on how Knowledge Graph entities power AI search results.

Frequently Asked Questions

How long does it take to start getting cited by ChatGPT and Perplexity?

Timeline varies significantly based on your starting point. Professionals who already have a Knowledge Panel and Wikidata entry typically see measurable citation frequency improvements within 60-90 days of content optimization. Those starting from zero entity recognition should expect 3-6 months to build the entity foundation, then 60-90 more days for content optimization to compound. Perplexity responds fastest because it’s real-time. ChatGPT’s training data has periodic updates, so changes take longer to register there.

Does publishing on LinkedIn help get cited by AI?

Indirectly, yes. LinkedIn posts aren’t typically cited as primary sources in AI responses, but strong LinkedIn presence contributes to your entity corroboration profile (Tier 3 signal), which strengthens entity recognition. More directly, LinkedIn articles and newsletters can appear in Perplexity’s real-time retrieval. The highest-impact LinkedIn action isn’t publishing more posts. It’s maintaining a complete, consistent profile that reinforces your entity home’s information.

Can I pay to get cited by ChatGPT or Perplexity?

No. There’s no paid citation pathway on ChatGPT, Perplexity, or Google AI Overviews. Citations are earned through entity signals and content quality. Google does offer advertising placements near AI Overviews, but those are ads — not citations. The citation itself, appearing as a named authority in an AI-generated response, requires organic entity building and content optimization. There are no shortcuts here.

What’s the difference between being cited anonymously vs. by name?

Anonymous citation means the AI extracted content from your page but didn’t name you. It might say “according to [source]” with a link, or paraphrase without any attribution at all. Named citation means the AI explicitly identifies you: “According to [Your Name], [your claim]…” That named citation requires entity recognition. The AI must have you in its model as a recognized Person entity. This is why Knowledge Graph optimization is a prerequisite for personal thought leadership visibility in AI. It’s not just a search tactic. It’s how you get credit for your own expertise.

How do I track whether AI is citing me?

Set up a weekly monitoring routine. Ask ChatGPT and Perplexity 5-10 key questions in your domain and note whether you’re cited. Use GA4 to track utm_source=chatgpt.com referral traffic (available since June 2025). For Google AI Overviews, search your target keywords in Google and observe whether your content or name appears in the AI-generated section at the top. Tools like DataForSEO and Conductor’s GEO benchmarks platform offer AI citation tracking at scale if you want to systematize it.

Conclusion

You know what separates the experts AI cites from the ones it ignores?

Not credentials. Not content volume. Not how good they are at their job.

It’s whether AI has been taught to recognize them.

My client with 14 years of experience, two books, and a keynote history wasn’t being penalized for anything she did wrong. She was invisible because no one had built the signals that teach AI systems to associate her name with her domain. That’s entirely fixable. And the fix is systematic, not random.

Start with the entity layer: Wikidata entry, Person schema, corroborating sources. Then format your content for extraction: answer capsules in every H2, FAQ schema on every article, dense introductions. Then track your citation profile and iterate.

Every day you don’t do this, someone with fewer credentials and better entity signals gets the recommendation instead.

Start with the entity layer. The rest compounds from there.

For the personal brand-specific application of these principles, see our guide to AEO for personal brands and thought leaders.