Solutions
PR CitationPlacements on 500+ scored crypto outlets AI actually cites AI CitationTrack how 5 AI engines cite you. 100-point audit, weights published ContentCitation-engineered articles. Humanized, fact-checked, tracked
Resources
LearnWhat AEO is and how AI search works for crypto, from scratch MethodologyEvery weight and formula behind the score, in the open Case studiesMeasured client results. Dated, reproducible, no vanity metrics ComparisonsciteOS vs Profound, Coinbound, and Cision, honestly ?FAQHow crypto citations work across AI and PR
More
Pricing Sign in Get my audit →
Learn / Where AI gets its crypto answers
Original data · 4,969 citations

Where AI gets its crypto answers: 9 sources, ranked by citation share

Every AI answer about crypto is assembled from somewhere. We pulled 4,969 citation events from 10 audited crypto brands across five engines, then read the pages behind 1,357 of them by hand, and tagged where each citation actually pointed. This is the ranked list, the share each source holds, and what it costs to get onto them.

By Sagar Saxena · Founder, Emergence Media · Last updated: July 29, 2026
The short answer
Where AI gets its crypto answers · the verdict

Nowhere concentrated. Across 4,969 citations the engines pulled from 979 different source domains, and 72.6% of all citation volume came from the long tail outside any named category. Your own domain is the single largest identifiable source at 10.75%, and that is still the smallest lever you have, because 89% of citations point somewhere you do not control. The winning move is not to publish more. It is to get listed on the pages that already get cited.

979
distinct source domains behind 4,969 citations. There is no shortlist to buy your way onto
72.6%
of citation volume comes from the long tail, outside every category below
46
citations earned by one single CoinGecko listicle. One page, one line, 46 citations
The list

The 9 sources, ranked by citation share

Share of total citation volume, not share of prompts. A source with 5% share supplied 5% of every citation the engines produced across the whole corpus. Ranked highest to lowest.

Citation share by source type · n = 4,969 citation events · 10 crypto brands · 5 engines · July 2026
#SourceShare of citationsWhat it costs to get on
1The long tail (everything uncategorised)72.60%Unbuyable. Earned by being a known entity
2Your own domain10.75%Free, and capped. You cannot grow past your own site
3Aggregators (CoinGecko, CMC, DefiLlama)5.17%Listing application, usually free to low four figures
4YouTube4.41%Creator fees, or your own channel
5Tier-2 crypto press2.88%From roughly $800 a placement
6Other community (Medium, Quora, Trustpilot)1.53%Free, slow, unreliable to force
7Reddit1.47%Free, and hostile to anything that looks bought
8Tier-1 crypto press1.13%From roughly $4,000 a placement
9X / Twitter0.06%Free, and almost never cited

Two rows deserve to be read together. Tier-1 crypto press holds 1.13% of citations. Tier-2 holds 2.88%. The cheaper tier was cited 2.55 times more often, at roughly a fifth of the price. In our corpus CoinDesk, Cointelegraph and Decrypt returned 10 citations between them. CryptoSlate, alone, returned 69.

That is not an argument against Tier-1. A Tier-1 placement still carries weight the volume numbers do not capture, and our published authority list is 1.9% of all source domains while carrying 11.77% of all citations, or 6.2 times its own weight. The point is narrower: you are almost certainly overpaying for volume and underpaying for the pages that actually get quoted.

Source by source

What each source is actually worth

1

The long tail · 72.60%

Nearly three quarters of citation volume comes from domains too scattered to categorise. No blog network, no vendor and no PR budget covers it. It is the residue of being a recognised entity across the open web, which is why entity presence sits inside the authority signal in our score rather than beside it. You do not buy this. You accumulate it.

2

Your own domain · 10.75%

The largest single identifiable source, and the one every project over-invests in. It has a hard ceiling: an engine composing a comparison answer will not build it out of one brand's own marketing. Publish, but treat your site as the thing that confirms a claim made elsewhere rather than the thing that makes it.

3

Aggregators · 5.17%

CoinGecko, CoinMarketCap and DefiLlama. Cheap, structured, and the highest ratio of effort to citation in the entire dataset. One CoinGecko listicle earned 46 citations on its own. If your listing is incomplete or your category tags are wrong, fix that before you commission a single article.

4

YouTube · 4.41%

The most cited external source, and the one with no incumbent. 166 different videos shared 219 citations, and the single most cited video earned 7. Compare that with the one CoinGecko page on 46. Aggregators are a fortress with a few doors. YouTube has no door at all: there is no video to outrank, because nothing dominates.

5

Tier-2 crypto press · 2.88%

Cited 2.55 times more often than Tier-1 at roughly a fifth of the cost. The reason is unglamorous: Tier-2 outlets publish more listicles and roundups, and listicles are what the engines quote. You are not buying prestige here, you are buying inclusion in comparison pages.

6

Other community · 1.53%

Medium, Quora and Trustpilot. Small, but disproportionately present in due-diligence answers where a buyer asks whether a brand is safe or legitimate. Worth monitoring, not worth a campaign.

7

Reddit · 1.47%

Ranked 9th of 9 by volume, against an industry consensus that treats it as a top channel. Reddit matters, but it matters as an existence signal that corroborates your brand, not as a citation source. Anyone quoting a large Reddit citation percentage is almost certainly reporting rate of appearance in a handful of prompts and calling it share of volume. Those are different measurements.

8

Tier-1 crypto press · 1.13%

The most expensive row on the table and the second-least cited. Buy it for the trust transfer and the human reader, and know that the direct AI citation return is small. Our outlet scoring exists precisely so this decision is made on measured signals rather than masthead recognition.

9

X / Twitter · 0.06%

Effectively zero. Three citations in every five thousand. X is where crypto marketing spends most of its attention and where AI engines look least. Post there for humans. Do not expect an engine to quote it.

Page format

One page format ran away with it

We hand-tagged the archetype of every page behind the 1,357 traceable citations. One format took 48% of them.

Citations by page archetype · hand-tagged, n = 1,357 traceable citations across 168 pages
ArchetypeCitationsPagesAvg per page
Listicle657798.3
Official brand page224336.8
Alternatives page98175.8
Brand review91146.5
Data directory87155.8
Explainer72107.2

Listicles took 657 of 1,357 traceable citations from 79 pages, the highest total and the highest average per page. Everything else is a rounding error beside it.

Then we tagged how each page treated the brand, which produced the least intuitive result in the study.

Average citations per page, by how the page treats the brand
How the page treats your brandAvg citations per page
Mentioned in passing8.3
Ranked in a list7.7
Absent entirely7.5
Subject of the page6.6
Compared against5.4

A page that mentions you in passing outperforms a page written about you. A page that does not mention you at all still outperforms your own brand page. The engine is not looking for your pitch. It is looking for a page that compares.

Which reframes the whole budget question. One line inside somebody else's ranked list beats a whole article about you, and it costs a fraction as much.

On-page elements

What to put on the page, with the measured lift

Four elements showed a repeatable citation lift against pages without them. The scale starts at 1.00×, so 1.16× means 16% more citations for an otherwise comparable page.

Citation lift by on-page element · scale starts at 1.00×
ElementCitation liftWhy it works
Real FAQ block1.16×Question-shaped text matches question-shaped prompts
Scored rating1.14×A number is extractable. An adjective is not
Explicit verdict1.13×The engine can lift a conclusion without composing one
Comparison table1.10×Rows and columns survive being parsed out of context

Individually minor. They stack, and none of it is expensive. A page carrying all four is running at roughly 1.6× the citation rate of the same page without them, for an afternoon of work.

Note what is missing from that list: word count, publishing cadence, keyword density. None of them separated cited pages from uncited ones in our corpus.

Consistency check

This is not one outlier brand

Ten brands across six verticals: centralised exchanges, a payments card, a lending protocol, a DePIN network and a tokenization platform. The shape holds across all of them.

Per-brand citation distribution · all 10 audited brands
BrandCitationsOwn domainYouTubeRedditAggregatorsSource domains
Brickken58217.5%1.4%0.2%2.2%185
Bitget55415.5%7.0%1.8%2.5%174
WhiteBIT54110.0%4.3%0.9%4.4%177
KuCoin52812.7%6.8%2.3%5.3%157
UPay5004.8%8.0%0.8%4.6%170
Maple Finance4797.7%0.8%1.0%4.0%198
Nexo4769.2%5.9%3.6%6.3%158
io.net43915.3%5.2%1.4%5.7%146
BetFury43812.1%1.4%2.3%1.8%149
1win4320.0%2.8%0.7%0.9%197

Every brand pulled from between 146 and 198 distinct source domains. Not one broke 18% on its own domain. 1win earned 432 citations with zero from a domain it owns, which is the cleanest demonstration in the set that owned content is optional and third-party presence is not.

The moves

Three moves, in order of return

1

Fix your aggregator listings this week

CoinGecko, CoinMarketCap, DefiLlama and every vertical directory in your category. Complete profile, correct category tags, current description. 5.17% of all citations, near-zero cost, and a single listicle on one of these earned 46 citations by itself. Nothing else in this article has that ratio.

2

Get into other people's ranked lists

Not articles about you. Lists that include you. Listicles took 48% of traceable citations, and a passing mention outperforms being the subject of the page. Find the "best X for Y" pages already ranking in your category and get added to them. This is the single highest-leverage line item in a crypto marketing budget, and almost nobody runs it deliberately.

3

Rebuild your own pages around the four elements

Real FAQ block, a scored rating, an explicit verdict, a comparison table. 1.16×, 1.14×, 1.13× and 1.10×, and they stack. Your own domain is capped at roughly 10% of your citations no matter what you do, so make that 10% count rather than trying to grow it.

Want your brand's version of this table rather than the industry average? The free scan runs 20 buyer prompts across 5 engines and returns your own citation split in about a minute, no card.

Run my free scan →
FAQ

Questions we get about this data

Is Reddit really only 1.47%?

By share of citation volume, yes, and it ranks 9th of 9 categories. The much larger Reddit figures circulating in AEO marketing are usually a rate of appearance across a small set of due-diligence prompts, presented as share of volume. Both are real measurements; they are not the same measurement. Reddit still matters as a corroborating existence signal, which is why entity presence carries 35% of the authority component in our score.

Why is Tier-1 press so low if it is so expensive?

Because volume and weight are different things. Tier-1 supplied 1.13% of citations while Tier-2 supplied 2.88%, largely because Tier-2 outlets publish more listicles and roundups. But our published authority list is 1.9% of source domains carrying 11.77% of citations, 6.2 times its own weight. Tier-1 buys trust and human readers. It does not buy citation volume.

Can I just publish more on my own site?

Only up to a point. Owned domains supplied 10.75% of citations across the corpus and no audited brand exceeded 17.5%. An engine composing a comparison answer will not assemble it out of one brand's own marketing. Publish to confirm claims made elsewhere, not to make them.

How was the archetype tagging done?

By hand. We opened the pages behind 1,357 of the 4,969 citations and tagged each one for archetype, brand treatment and on-page elements. The remaining citations pointed at pages that were unreachable, paywalled or duplicated, and were excluded from the archetype analysis rather than guessed at.

Does this hold outside crypto?

Unknown, and we will not claim it does. Every brand in the corpus is a crypto brand and every prompt is a crypto buyer prompt. The archetype and on-page findings are plausibly general. The source-share table is almost certainly not, because aggregators like CoinGecko have no equivalent in most industries.

Will my numbers look like these?

Directionally yes, precisely no. Every brand in the study pulled from a different set of 146 to 198 source domains. The shape held; the specifics did not. That is the reason to measure your own rather than plan against an industry average.

SS
Sagar Saxena · Founder, Emergence Media

Growth marketer working in Web3 since its early days; has helped 75+ crypto and Web3 projects reach their target audiences. Runs Emergence Media and builds citeOS. This study is drawn from citeOS audit data, and the full methodology behind every number is published in the open.

Find out if Perplexity recommends you.

Five engines, twenty buyer prompts, 100 observation points. Published weights, reproducible citations, your baseline in about a minute. No card, no signup.

Scan my brand free →