GEO for crypto payments

Across 39,948 citation events covering 72 crypto brands between April and August 2026, payment and onramp pages account for 4,284 citations across 937 distinct pages. The top of that list is where the concentration sits.

Those nine carry 776 citations. The five vendor-published pages account for 459 of them, or 59%.

QuickNode and Alchemy are counted as independent because they sell infrastructure rather than competing gateways. On a looser definition the vendor share is seven of nine. The conservative reading is used throughout.

Cards ties payments on page count, but one issuer holds two of those five pages, so it is four distinct vendors rather than five. Nexo appears twice with two separate URLs.

Cards matches payments on page count and does not match on volume. In wallets the top two pages are both independent and the first vendor page sits third. Payments is the only vertical measured where a vendor page takes the top slot outright.

Because retrieval matches the shape of the question, not the publisher's incentives.

A prompt like "best crypto payment gateway for a business" wants a ranked list of named products with criteria attached. A gateway that publishes exactly that page has produced the right object. Nothing in the retrieval step establishes whether the author competes in the category.

This matches the wider corpus. Ranked lists are 33% of all citations recorded, rising to 48% among the most-cited pages read by hand, and being the subject of a page earns 6.22 citations against 3.78 for being one name inside someone else's list. Those figures come from the 39,948-citation study, with the source-by-source breakdown in where AI gets its crypto answers.

Format does more work than neutrality.

That describes current engine behaviour. It is not an endorsement and it is not stable. An engine that starts weighting publisher independence reprices this overnight, so treat it as a position that can move, not a moat.

• Publish the category page. Five companies in your category already have and are being cited for it. Waiting for an independent outlet to rank you is waiting behind competitors who did not wait.
• Disclose, or it stops working. A list that ranks the publisher first and concedes nothing reads as an advertisement, and being caught doing that costs more than the citations are worth. Name real criteria, apply them evenly, and say plainly that you are one of the companies being ranked. The pages that hold up are the ones a competitor could read without objecting.
• Do not mistake it for a strategy. Across the full corpus 92.22% of citations point at pages the brand does not control and own-domain pages account for 7.78%. Your category page sits inside that 7.78%.
• Keep widening coverage. Citations scale with the number of pages mentioning a brand at roughly 1.48 per page, r = 0.80, with a coefficient of variation of 9.5% across brands of very different sizes. No tactic tested beat being written about in more places. The mechanism is in how AI picks crypto sources.

The vendor and independent labels are a judgement call made by reading each publisher, not a measured property of the data. Reasonable people would move QuickNode and Alchemy.

Counts are per page across all 72 audited brands, April to August 2026. Five months is a short window, engine behaviour changes, and this source map is crypto only.

Citation counts are not traffic and not revenue. No causal claim connects a cited page to a transaction.

Engines are non-deterministic, so running one of these prompts and getting a different source list does not contradict any of this. Every figure is a property of the pool across 39,948 events, not a prediction about a single answer. The scoring weights and the full method are published at citeos.io/methodology.