Most cited domains in ChatGPT
Across millions of ChatGPT answers, a small set of domains shows up as citations far more often than the rest. The list is predictable: Wikipedia and Reddit at the top of nearly every category, GitHub and Stack Overflow for technical queries, plus the dominant review and trade publications per industry.
The domains that appear across nearly every category
Five sources dominate ChatGPT citations regardless of topic. Wikipedia shows up on roughly 60 to 80% of factual queries, the highest base-rate citation source by a wide margin. Reddit appears on 30 to 50% of opinion or recommendation queries, especially anything involving comparison or buyer experience. GitHub dominates technical and developer-tool queries. Stack Overflow dominates coding and engineering troubleshooting. YouTube appears across product reviews and how-to queries.
These five together account for a significant share of citations on most queries. The remaining citation slots go to category-specific sources.
Category-specific patterns
The non-generalist citation slots are predictable per industry. SaaS and B2B software: G2, Capterra, TrustRadius, Gartner, plus the top 5 trade publications in the niche (think TechCrunch for general tech, Search Engine Land for SEO, Reuters for finance). Healthcare: NIH, Mayo Clinic, Cleveland Clinic, Healthline, WebMD, plus specialist journals. Finance: NerdWallet, Investopedia, Bankrate, plus the SEC and major banks' own sites. Legal services: state bar association pages, Justia, Avvo, plus the firm's own pages when they have substantive content. Local services: Yelp, Google Maps reviews aggregator data, local newspaper sites, plus the business's own site when properly schema-marked.
The pattern: ChatGPT cites the authoritative aggregator first and the individual brand second. Earning placement on the aggregator usually drives more downstream citation than improving your own site directly.
Why these specific domains dominate
Three structural reasons. (1) Schema and crawlability: the dominant sources almost universally have FAQPage and Article schema, server-rendered HTML, and explicit AI-crawler allowlists. (2) Topical authority: ChatGPT's re-ranker weights per-topic authority more than overall domain authority. A specialist site on a narrow topic outranks a generalist with higher classical metrics. (3) Extractive structure: pages that answer specific questions in 2 to 3 sentence blocks get cited at much higher rates than essay-form content because they slot into ChatGPT's answer generation more cleanly.
Domains that lose: pages with marketing fluff, slow load times, JS-only rendering, or no structured data. ChatGPT's re-ranker actively de-weights these.
How to use this for your AEO strategy
Two practical moves. (1) Audit your category's actual top cited domains rather than working from the generic list above. Run 40 to 100 buyer-intent queries in your category through ChatGPT, log the citation domains, count them. Your top 10 is the real PR target list, often surprising compared to traditional SEO competitor lists. (2) Pitch to that target list, not just earn classical backlinks. Earned mentions on aggregators that already get cited in your category compound into your own citation rate within 2 to 4 weeks, faster than direct on-site work because you inherit the aggregator's existing AI trust signal.
Our public industry rankings show the top cited domains per category for the industries we've audited, refreshed monthly.
Related concepts
Frequently asked
Is Wikipedia really the most cited domain across the board?
Yes, by a wide margin. Across nearly every factual or definitional query, Wikipedia appears as a citation at base rates of 60 to 80%. It is the closest thing to a universal source in ChatGPT's citation graph.
Why does Reddit show up so often?
ChatGPT explicitly leans on Reddit for opinion, comparison, and user-experience signals. For 'best X' queries Reddit is often cited even when more authoritative sources exist, because Reddit threads carry the lived-experience perspective the answer needs.
Do the top cited domains vary by country?
Somewhat, yes. UK queries cite UK trade publications (Drapers, The Grocer, Computer Weekly) more than US equivalents. Australian queries pull from Australian review sites. The encyclopedic sources (Wikipedia, GitHub, Stack Overflow) stay constant globally.
Should I try to get listed on G2 and Capterra?
For SaaS and B2B software, almost always yes. Both consistently appear in ChatGPT citations for any commercial-intent SaaS query. Even a baseline profile with a handful of reviews is enough to start surfacing in the AI's source list.
How do I find the most cited domains in my specific category?
Run a representative query set (40 to 100 buyer-intent questions in your category) through ChatGPT and tally the citation domains. The top 5 to 10 are your real PR target list. Our public industry rankings show this data per category for industries we have already audited.