We pulled 88,975 citation events from 81 audited crypto brands across the 5 AI engines buyers actually use, then read the pages behind 1,357 of those citations by hand, to answer one question: what does it really take to get cited by AI? The findings break a lot of marketing intuition. Here is the data, free.
Live update, 30 Oct 2026. Counts on this page track the live citeOS corpus: 88,975 citation events from 81 audited brands across 7,432 source domains, June to October 2026. The percentages come from the original April to August 2026 run of 39,948 citations and still hold on the live data: off the brand's own domain 91.47% live against 92.22% here, YouTube 4.61% against 4.49%, and 107 domains carrying half of all citations against 99.
n = 88,975 citation events · 81 audited crypto brands · 7,432 source domains · pages behind 1,357 citations read by hand
No vibes, no vanity metrics. Every claim in the report is tied to a number you can act on.
You will get the full cite-rate breakdown by outlet tier, the engine-disagreement data, the methodology behind every number, and a plain-English playbook for where to spend your next marketing dollar.
And the horizontal AEO tools were built for SaaS: they don't rank crypto outlets, understand a token launch, or know a DeFi answer shouldn't cite a gaming blog. So we measured crypto on its own terms. This report is the first cut of that data, and the tool that measures your own version of it goes live this week.
Want your brand's version of this data, not the industry average?
Run my free scan →The report is the industry picture. The free scan is your picture: 5 engines, 20 buyer prompts, your real citation gap: in about a minute, no card.
Get the free report →