Stablecoin Spend Report — Q3 2026

This is the first Stablecoin Spend Report, which means it is a baseline rather than a trend. There is no previous quarter to compare against, and we are not going to manufacture one. Every figure below describes the directory as it stood at the end of the period; from the next edition onward these numbers become the thing new numbers are measured against.

That constraint is worth stating up front because the genre invites the opposite. Quarterly adoption reports in this sector routinely open with a growth percentage derived from a dataset that did not exist a quarter earlier. Ours did not either. So: no growth claims this time.

The headline finding is a disclosure problem

The most useful thing in this dataset is not which stablecoin leads. It is how much merchants decline to say.

Close to half of the merchants that accept stablecoins — 47 of the 104 in this snapshot — do not publish which blockchain network they settle on. A smaller group confirm that they accept cryptocurrency without naming a single asset — the checkout will tell you, and until then you are guessing.

This is not a research gap on our side. In almost every case we reached the merchant’s own payment documentation and it simply does not contain the information. The merchant has integrated a gateway, the gateway maintains the asset list, and the merchant has no particular incentive to mirror it.

For a buyer this is the whole problem in one sentence. Sending USDT to the wrong chain destroys the money, irreversibly, and the merchant page you are reading before checkout frequently does not tell you which chain is correct. Everything else in this report is downstream of that.

USDC leads USDT, and regulation is why

USD Coin appears at more merchants than Tether does, which inverts the ranking you would predict from market capitalisation. Tether is by a wide margin the larger asset by supply and by on-chain transaction volume. It is not the one merchants put on their checkouts.

The mechanism is visible in the individual records rather than in the aggregate. Under the EU’s Markets in Crypto-Assets regulation, a stablecoin offered to EU customers requires its issuer to hold an e-money authorisation. Tether did not obtain one. European payment processors withdrew USDT for EU-established merchants through late 2024 and 2025, and the merchants followed.

Two records in this directory document the change from the merchant’s side rather than the regulator’s. Cherry Servers published a migration notice and enforced a hard stop on USDT on 1 April 2025, directing customers to USDC. HostSailor, also EU-established and also on CoinGate, simply cannot offer Tether.

The consequence for anyone holding USDT and shopping in Europe is concrete: expect to swap. The consequence for the market is that a regulatory decision, not a merchant preference, is currently shaping which dollar token is spendable where.

Acceptance is gateway acceptance

Almost no merchant in this dataset takes a direct wallet transfer. The overwhelming majority route through a payment gateway, and a meaningful number do not name which one.

This matters more than it first appears, because it means merchant “acceptance” is mostly a pass-through. The merchant chooses a processor once. The processor decides, and revises, which assets appear at checkout. When the list changes the merchant usually does not announce it — frequently does not know — and the customer discovers it at the payment step.

It is why so many records here carry the status payment availability may vary rather than a fixed coin list. That status is not hedging. It is the accurate description of a two-party arrangement where the party you are buying from does not control the answer.

The corollary for merchants reading this: if you accept stablecoins and you publish your asset list and your settlement networks on your own domain, you are doing something a large majority of your peers are not, and it is the single cheapest thing you can do to reduce failed payments.

The category distribution has not moved off infrastructure

Hosting, VPS, domains, VPNs and proxies dominate. Physical retail is close to absent. Travel and gift cards form a second cluster, and everything else is thin.

The reasons are structural rather than incidental. Infrastructure businesses sell to a technically sophisticated audience that already holds crypto; they bill small recurring amounts where card processing overhead is proportionally high; they have low chargeback exposure because the product is delivered instantly and disputes are rare; and a good number of them serve customers who actively prefer not to route a card through their own name.

None of those conditions hold for a supermarket.

Anyone arguing that stablecoins are becoming a general-purpose consumer payment rail should be asked to account for this shape, because it has been stable for years and this quarter’s data does not disturb it. What has changed at the margin is the arrival of large conservative brands through account-based providers — Sony’s Singapore store, Emirates in the UAE — which is a genuinely different pattern from the wallet-to-gateway model, and worth watching.

What we withdrew

Research is not only additive, and a report that only counts what was added is not measuring honestly.

Menufy, a US restaurant ordering platform, accepted four dollar stablecoins through BitPay from 2020. It withdrew cryptocurrency entirely in July 2025 and says so in a dated note on its own site. That record is now the clearest example in the directory of the everyday-consumer-spending use case being tried at scale and abandoned.

Several other candidates researched this quarter were rejected before publication because the evidence did not support the claim — merchants that turned out to accept Bitcoin only, merchants whose crypto option had quietly disappeared, and merchants where every source traced back to a directory copying another directory. The rejections are not visible on the site, which is the point, but they are a larger number than the additions.

What the next report can measure that this one cannot

With a stored baseline, the Q4 edition will be able to report actual movement: merchants added, acceptance gained and lost, networks appearing and disappearing, and gateway switches. Those will be computed from the same snapshots rather than asserted.

Three things we will be watching specifically:

Whether USDT recovers in Europe. Tether has signalled intent to serve the EU through a MiCA-compliant vehicle. If that lands, some of the merchants that dropped it may add it back, and the changelog will show it.

Whether Shopify’s USDC integration produces visible merchant-level acceptance. Shopify Payments can now settle USDC on Base across a large merchant base, but a toggle being available is not the same as stores turning it on, and a store that has enabled it rarely says so on its own site. This may be a category of acceptance our methodology structurally cannot see, which would be an important limitation to name rather than paper over.

Whether the disclosure gap narrows. It is the metric we would most like to see move, and the one we have least reason to expect will.

Method

Every figure is computed from the merchant database at build time by the same code that produces the Stablecoin Spend Index, then frozen into a stored snapshot bound to this report. Nothing is typed in by hand.

The dataset counts merchants, not transactions. It covers businesses we could verify from primary sources, which biases towards those documenting payment options in English. It is a sample and not a census, and the full limitations are set out on the Index page and in our verification methodology.