Why ChatGPT, Perplexity and Google AI Overviews Cite Different Sites

Only about 11% of domains are cited by both ChatGPT and Perplexity for similar prompts, according to an analysis of 680 million AI citations. ChatGPT leans heavily on Wikipedia, Perplexity favours Reddit, and Google's AI Overviews lean on YouTube. If you have only optimised for one of these platforms, you are likely invisible on the other two.

What changed?

A Profound analysis of 680 million citations found only around 11% of domains are cited by both ChatGPT and Perplexity for similar prompts. The gap between platforms is not marginal, it is structural.

ChatGPT leans on Wikipedia for 47.9% of its top-10 citations, Perplexity leans on Reddit at 46.7%, and Google AI Overviews leans on YouTube at 23.3%, according to the same Leapd Blog write-up. These are not close scores. Each engine has effectively built its own citation ecosystem, and the overlap between them is small enough that treating "AI search" as a single channel misses most of what is actually happening.

A separate write-up of the same underlying study, from Digital Applied, confirms the headline figure: across ChatGPT, Google AI Overviews and Perplexity, only 11% of domains are cited by both ChatGPT and Perplexity, with each platform operating on fundamentally different citation logic.

This matters because most GEO advice still treats "getting cited by AI" as one job. The data suggests it is at least three separate jobs, each with its own source preferences, content formats and trust signals.

Why does this matter for your business?

If you have spent the past few months building a single GEO playbook and applying it everywhere, you may be optimising for the wrong platform without realising it.

Take review and comparison content. G2 is the most cited software review platform on ChatGPT, Perplexity and Google's AI Overviews, according to Position Digital's roundup of AI SEO statistics. That is one rare point of agreement across engines, which is exactly why third-party review sites matter so much for B2B brands right now. Outside review-heavy categories, though, the picture fragments quickly.

LinkedIn is the most-cited domain for professional queries in AI Overviews, AI Mode, ChatGPT, Microsoft Copilot and Perplexity, per Position Digital. That is another shared signal, useful if you publish B2B thought leadership. For general informational and community-driven queries, though, the platforms diverge sharply: YouTube and Reddit combined account for 78.2% of AI social media citations, with YouTube cited in 31.8% of cases and Reddit in 46.4%, according to the same source.

The practical implication is blunt. A brand with a strong Wikipedia presence and structured entity data may do well in ChatGPT, but if it has no Reddit footprint, it risks being absent from Perplexity's answers on the exact same topic. A brand with strong YouTube content may show up in Google's AI Overviews while being invisible in ChatGPT's more encyclopaedic citation pattern.

There is also a visibility trap worth flagging. Most LLM citations are "ghost citations," where the domain gets a source link but the brand name isn't mentioned in the answer text, and 61.7% of LLM citations fall into this category, according to Position Digital. Even when you are cited, you may not be named. That compounds the platform-fragmentation problem: you could be technically present in an AI answer's source list on one engine while being both unnamed and completely absent from another.

Ranking position still plays a role, but less cleanly than in classic SEO. Ahrefs data cited by Position Digital shows that 47% of AI Overview citations in 2025 came from pages ranking below position 5 in organic search, while SeoClarity's analysis of 432,000 keywords found that 97% of AI Overviews cite at least one source from the top 20 organic results, not just the top 10. Ranking in the top 20 gets you into contention, but it does not guarantee which platform, if any, will actually pick you up.

The commercial stakes are real too. Pages cited in AI Overviews earn 35% more organic clicks than non-cited competitors on the same results page, and visitors arriving from Perplexity convert at roughly 11 times the rate of traditional organic search traffic, per Position Digital. Missing out on citations on even one platform is not a cosmetic problem, it is a lost revenue channel.

What should you do now?

Stop treating GEO as one project with one checklist. Instead, build separate, platform-aware tactics.

For ChatGPT, invest in entity clarity and third-party encyclopaedic presence. Given its heavy reliance on Wikipedia-style sourcing, make sure your brand, founders and products have clean, factual, well-referenced public records. Wikipedia notability is a long game, but structured company facts pages, press coverage and consistent public data all feed the same trust signals ChatGPT appears to favour.

For Perplexity, build a genuine community presence. With Reddit accounting for nearly half of Perplexity's top citations, participating authentically in relevant subreddits, and being the kind of brand real users discuss unprompted, matters more here than almost anywhere else in GEO.

For Google AI Overviews, invest in video. YouTube's outsized role in Google's AI citations means explainer videos, product demos and tutorial content are not optional extras, they are a direct route into Google's answer engine.

Across all platforms, get listed and reviewed on G2 and equivalent sites if you sell software or B2B services, and keep LinkedIn content active if you serve professional audiences. These are the rare cross-platform wins the data supports.

Finally, check whether you are showing up at all, and whether you are being named when you do. Ghost citations mean a source link alone is not proof of real visibility, so it's worth running your brand through Sited's free AI visibility audit at https://sited.online to see how you're actually described, or omitted, across these different AI platforms, which is the only way to know if your strategy is working per engine rather than in aggregate.

Frequently asked questions

Does ranking well on Google guarantee AI citations?

Not on its own. Ahrefs found 47% of AI Overview citations in 2025 came from pages ranking below position 5, while SeoClarity found 97% of citations came from somewhere in the top 20 organic results. Top 20 gets you into the running, it does not decide which platform, if any, picks you.

Why would a brand be cited by ChatGPT but not Perplexity?

Because only about 11% of domains are cited by both ChatGPT and Perplexity for similar prompts, with ChatGPT leaning on Wikipedia, Perplexity on Reddit, and Google AI Overviews on YouTube. Each platform's sourcing habits are different enough that presence on one does not transfer to another.

What is a "ghost citation" and why should I care?

It is when the domain gets a source link but the brand name isn't mentioned in the answer text. You may technically be a source without a user ever seeing your brand name, so link presence alone is not a reliable visibility metric.

Which platforms should B2B brands prioritise first?

Review platforms and professional networks show the most cross-platform agreement. G2 is the most cited software review platform on ChatGPT, Perplexity and Google's AI Overviews, and LinkedIn is the most-cited domain for professional queries across AI Overviews, AI Mode, ChatGPT, Microsoft Copilot and Perplexity. These are sensible starting points before you branch into platform-specific tactics.

Is it worth optimising for AI citations if it hurts click-through rates?

The data suggests otherwise. Pages cited in AI Overviews earn 35% more organic clicks than non-cited competitors, and Perplexity visitors convert at roughly 11 times the rate of traditional organic traffic. Citation and traffic quality are not in conflict here, they reinforce each other.