Key takeaways
- ChatGPT, Google AI Overviews, and Perplexity source differently. Profound's analysis of 680 million citations found the biggest single source differs by platform: Wikipedia is 7.8% of ChatGPT's citations, Reddit is 6.6% of Perplexity's and 2.2% of AI Overviews'.
- Those are small shares of all citations, so most citations on every platform go to a long tail of other sites. The 47.9% and 46.7% figures often quoted are shares of each platform's top 10 sources, not of all citations.
- Optimizing for one platform does not optimize for all three. Check which assistants your buyers use before you weight the work.
- Reddit is the largest single cited source on two of the three platforms, but a small slice overall. Community presence is worth earning, not a substitute for clear pages of your own.
- This post covers the platform differences. The evidence on schema, the GEO research and on-page habits is in our complete guide to answer engine optimization.
What this post covers
Search "how to get cited by ChatGPT" and you'll find dozens of near-identical checklists. Almost none cites a real study. The full evidence base, including the peer-reviewed GEO paper, the Ahrefs schema experiment and what to change on the page, is in what is answer engine optimization. Here we cover the one question that guide treats briefly: do the major assistants cite the same sources, and what does that change about where you spend effort?
It depends which AI you're trying to reach
Profound analyzed 680 million citations across ChatGPT, Google AI Overviews, and Perplexity between August 2024 and June 2025, the largest dataset we found on this question (Profound, AI Platform Citation Patterns). The three platforms do not source information the same way, and the difference is large enough to change where you'd spend effort.
One detail is easy to misread. Profound publishes two sets of percentages: each source's share of all citations on a platform, and each source's share within that platform's top 10 most-cited sources. The headline figures you see quoted (Wikipedia at 47.9% of ChatGPT, Reddit at 46.7% of Perplexity) are the second kind. They describe how the leaders compare with each other, not how much of all citations they capture.
Where ChatGPT's citations actually come from
Share of all ChatGPT citations in Profound's dataset: its two largest sources vs the long tail.
7.8% Wikipedia
1.8% Reddit
90.4% All other sources
Profound, AI Platform Citation Patterns, 680 million citations, Aug 2024-Jun 2025 (Profound sells AI visibility software)
Where Perplexity's citations actually come from
Share of all Perplexity citations in Profound's dataset: its two largest sources vs the long tail.
6.6% Reddit
2% YouTube
91.4% All other sources
Profound, AI Platform Citation Patterns, 680 million citations, Aug 2024-Jun 2025 (Profound sells AI visibility software)
Where Google AI Overviews' citations actually come from
Share of all Google AI Overviews citations in Profound's dataset: its two largest sources vs the long tail.
2.2% Reddit
1.9% YouTube
95.9% All other sources
Profound, AI Platform Citation Patterns, 680 million citations, Aug 2024-Jun 2025 (Profound sells AI visibility software)
The leading source differs by platform. Wikipedia is the largest single source for ChatGPT (7.8% of its citations), while Reddit is the largest for Perplexity (6.6%) and for Google AI Overviews (2.2%). Wikipedia does not appear in Perplexity's top 10 at all, and is only 0.6% of AI Overviews citations. On the overall measure, most citations on every platform go to a long tail of other sites.
The practical read is more modest than "go win Reddit". None of the leading sources passes 8% of any platform's citations. Genuine, substantive participation in the community discussion happening in your category (real threads, real answers, not disguised marketing) is worth doing, but it is one source among hundreds. Profound sells AI visibility software, so treat its framing with the same scrutiny you would apply to any vendor's data.
The sources also differ by industry
Our own audits point the same way. In the law firm audit, the firms' own sites were 0.6% of cited sources and ranking directories did the work. In the insurance audit, Yelp was the most-cited source. In the marketing agency audit, a competing agency's site was the most-cited domain. These audits measure who is named and which sources are cited, not which tactic caused it, but they show that "which platform" and "which industry" both change where the answer is built. All twelve are on our AI visibility research page.
What to do with this
1. Know which platform your buyers actually use, and weight accordingly. If your audience leans ChatGPT, Wikipedia-adjacent authority and reference-grade clarity are worth attention. If they lean Perplexity or Google AI Overviews, genuine presence in community discussion deserves a place in the plan, though no single source passes 8% of any platform's citations.
2. Check which sources your category leans on. Run buyer questions that never contain your name and record the cited sources. Directories, review sites and competitor lists may matter more than any platform average.
3. Publish something with a number nobody else has. Original data is the one category of content a lower-authority site can use to compete regardless of platform, because nobody else can cite a number they haven't measured.
4. Do the on-page basics. A direct answer near the top, named sources and accurate markup are covered, with the evidence and its limits, in our AEO guide.
Run your own content through the tool below rather than assuming any of this applies to your category the way it applied to ours.
Content Strategy Score
Score your own content against this framework
Supplied by DeliBrief. Want this on your own blog articles?
Book a quick call and we'll implement it for youBook a free consultation and we will run realistic buyer questions against live assistants and show you who is named instead of you.
Book a consultationFAQ
Do ChatGPT, Google AI Overviews, and Perplexity cite the same kinds of sources?
No, and the difference is large enough to change strategy. Profound's analysis of 680 million citations found the largest single source differs by platform: Wikipedia is 7.8% of ChatGPT's citations, while Reddit is 6.6% of Perplexity's and 2.2% of Google AI Overviews'. Those are shares of all citations, so most citations on every platform go to a long tail of other sites. Within each platform's top 10 sources the skew looks starker (Wikipedia 47.9% of ChatGPT's top 10, Reddit 46.7% of Perplexity's and 21.0% of AI Overviews'), but that is a share of the leaders, not of all citations. Optimizing for one platform does not mean you're optimized for all three.
Should I focus on Reddit?
Not as a strategy on its own. Reddit is the largest single named source on Perplexity and Google AI Overviews in Profound's data, but only 6.6% and 2.2% of those platforms' citations. Genuine participation in the discussions in your category is worth doing. It is one source among hundreds, not a replacement for clear, citable pages of your own.
Does schema markup or the GEO research change which platform I should target?
Neither tells you which platform to weight. For what the evidence says about schema (no reliable citation uplift in Ahrefs' study of 1,885 pages) and about the Princeton GEO paper (up to roughly 40% visibility gains in a simulated engine, varying by domain), see our complete guide to answer engine optimization.
How do I find out which sources an assistant uses for my category?
Run realistic buyer questions that never contain your name through live assistants and record who is named and which sources are cited. Our free AI visibility checker does this for five queries, and the same method is behind our 12-industry research. The sources differ by industry as well as by platform: review sites for insurance, ranking directories for law firms, Wikipedia and patent records for manufacturers.
Related reading
- What is answer engine optimization (AEO)? The complete guide
- Generative engine optimization (GEO): the complete guide
- AEO vs GEO vs SEO: what actually separates them
- AI visibility research: every audit in one place
- What 912 million blog posts actually show about backlinks and social shares
- Inbound marketing: what actually brings in buyers
Sources
- Profound, AI Platform Citation Patterns: How ChatGPT, Google AI Overviews, and Perplexity Source Information, June 2025, updated August 2025 (680 million citations)
- Aggarwal, Murahari, Rajpurohit, Kalyan, Narasimhan & Deshpande, GEO: Generative Engine Optimization, KDD 2024
- Ahrefs, We Tracked 1,885 Pages Adding Schema. AI Citations Barely Moved.
- Deligatr, AI visibility research across 12 industries
