44% of Every Source AI Cites Comes From 1% of Websites

Paris Childress
June 16, 2026

GEOforge Research · AI Citations

We analysed 1.35 million citations. AI doesn't read the web. It reads the same few dozen sites, over and over.

1.35M
citations analysed
12,317
distinct domains cited
44.5%
of citations from the top 1% of domains
28.9%
from just the top 50 domains

"Get cited by AI" has become the GEO mantra. But cited where? The data shows the field is far narrower, and far more winnable, than "the whole web" suggests.

What a citation is, and why it decides who AI trusts

A citation is a source the model leans on to build its answer: the pages and domains it pulls facts and names from when it responds to a query. Citations matter because they're upstream of mentions. If your brand shows up in the sources AI trusts, it shows up in the answer. If it doesn't, no amount of on-site optimisation will put it there. So the practical question is which sources actually carry that weight.

The answer is a steep power law

Across 1.35 million citation instances spanning 12,317 distinct domains, the distribution wasn't remotely even. The top 1% of domains accounted for 44.5% of everything cited. The top 50 domains alone supplied 28.9%. The top 10 carried nearly 11% on their own. A long tail of more than twelve thousand sites splits what's left.

Share of all AI citations by domain rank

Top 10 domains10.9%
Top 50 domains28.9%
Top 1% (~123 domains)44.5%

AI doesn't survey the open web for every answer. It returns, again and again, to a small set of sources it treats as authoritative. Get into that set and you're cited everywhere. Stay out of it and you're competing in a tail that barely registers.

Who are the recurring sources?

Some are universal. Wikipedia appeared in the citation set for 16 of the 21 brands we track. Reddit showed up for 14. Forbes for 11, Gartner for 8, G2 for 6. Others are category-specific heavyweights, like Semrush and Clutch in marketing, or the scientific and vendor-documentation domains that dominate technical fields. The mix shifts by industry, but the structure is always the same: a handful of sources do most of the work.

Cross-category sources AI keeps returning to
SourceBrands it was cited for
Wikipedia16 of 21
Reddit14 of 21
Forbes11 of 21
Gartner8 of 21
G26 of 21

Of 1.35 million sources ChatGPT cited, 44.5% came from the top 1% of domains. You don't need to be everywhere. You need to be in the few dozen places AI keeps going back to.

This is good news, if you act on it

A power law means the problem is bounded. "Earn citations across the web" is paralysing. "Earn your place in the 50 to 100 sources that drive nearly half of AI's citations in your category" is a project plan. Identify the recurring domains for your space, audit which ones already mention competitors and not you, and go earn presence there deliberately. That's a finite, prioritised list, not an ocean.

How to win the sources that matter

  • Map the domains AI actually cites in your category, ranked by frequency.
  • Find the recurring sources that mention competitors but not you.
  • Prioritise the universal ones first: Wikipedia presence, relevant subreddits, analyst and review platforms.
  • Earn the citation honestly, with content worth referencing, not link spam.
  • Re-measure citation share, not just mentions, so you can see the source base shift.

Tracking which sources AI trusts is table stakes. Getting into them is the work that moves your visibility.

Find the sources AI trusts in your category. Then earn them.

GEOforge maps the domains driving AI citations in your market, shows where competitors are cited and you aren't, and builds the presence to close it.

Book a GEO visibility audit →

Sources & method. All figures verified against the GEOforge measurement database on 16 June 2026. Corpus: 1,352,770 citation instances across 12,317 distinct domains, drawn from 51,391 classified citation source URLs across 21 tracked brands, measured April–June 2026. "Top 1%" is the highest-frequency 1% of domains by citation volume. Cross-category source counts reflect how many of the 21 brands each domain was cited for. Figures are ChatGPT-specific; cross-engine results may differ.

Paris Childress
CEO

Paris Childress is the CEO of Hop AI and creator of GEOforge, a platform that helps B2B brands get cited and recommended by AI assistants like ChatGPT, Perplexity, and Gemini. A former Google Country Manager and agency veteran with 20+ years in digital marketing, Paris is focused on helping brands win in the era of AI search.