Fair Payments Alliance · fairpayments.org
A published test for whether a fee on a merchant statement is honest. Six questions, four grades, one method — applied the same way to every processor.
Every merchant knows the feeling of reading a statement and not understanding a line on it. The industry relies on that. Fees are named to sound mandatory, disclosed in ways that satisfy a contract but not a person, and raised in increments small enough that nobody calls.
There is no shortage of complaint about this. What is missing is a line — a published, specific test that says what fair actually looks like, so that a fee can be measured against something other than how it feels.
This document is that test. It is written so that a merchant, a journalist, a regulator, or a processor's own compliance team can apply it and reach the same result. If we grade a fee and you think we got it wrong, the rubric is right here — show your work and we will publish the correction.
The Standard applies to any recurring or one-time charge that appears on a merchant's card processing statement, other than interchange and card network assessments, which are set by the card brands and are not the processor's to price.
The Standard does not grade whether a price is high. A processor is entitled to charge what the market will bear, and a merchant is entitled to pay it. What the Standard grades is whether the merchant could see it coming, understand what it is, and know what it is for. A $2,000 fee that is plainly named, fully disclosed, and honestly explained passes. A $10 fee that is dressed up as a regulatory requirement does not.
Expensive is a business decision. Deceptive is a different thing entirely. This Standard is only about the second one — which is why a processor can price aggressively and still earn a clean grade.
Each fee is put to six questions. Every question is answerable from two documents a merchant already has or can request: the merchant agreement and the statement, supplemented by whatever the processor has published itself.
Was it disclosed at signing?
Was the fee in the agreement the merchant actually signed — named, priced, and findable — or was it introduced later under a change-of-terms clause?
Passes when the fee and its amount appear in the executed agreement or its attached schedule. Fails when the only authority for it is a clause reserving the right to add fees at will.
Was proper notice given?
For any fee added or increased after signing: was the merchant told clearly, in advance, in a communication whose purpose was to tell them?
Passes with direct written notice, stating the amount and the effective date, far enough ahead that the merchant can act on it. Fails when the notice is a statement insert, a footer line, a help-desk page nobody was pointed to, or arrives after the charge.
Is it cost-justified?
Does the fee correspond to an actual cost the processor incurs on this merchant's behalf, or is it margin with an operational-sounding label?
Passes when the underlying cost is real and identifiable — a network assessment, a physical device, a service actually rendered. Fails when the same fee is charged in different amounts to different merchants for identical coverage with no stated basis, or when no underlying cost can be identified at all.
Is it honestly named?
Does the name of the fee, and the processor's own description of it, describe what it actually is?
This is the test the industry fails most often, and the one that does the most damage. A fee named for a regulator implies a regulator requires it. A fee named for compliance implies non-payment creates exposure. Fails whenever the name or description attributes the charge to an authority, mandate, or cost that does not exist. Naming a fee vaguely is a lesser problem than naming it falsely — but both are marked.
Can the merchant escape it?
Can the merchant decline the fee, decline the service it pays for, or leave without a penalty that exceeds the fee itself?
Passes when the service is optional and separable, or the merchant can exit on reasonable notice. Fails when the fee is bundled into an unavoidable package and the exit is gated by an early-termination charge, an equipment buyout, or an auto-renewing term.
Has it stayed stable?
Has the fee risen over time without a corresponding rise in the cost it represents?
Passes when increases track a documented cost change or are absent. Fails when the price has climbed materially while the described service has not changed — the pattern merchants recognize as the rate that quietly creeps.
Test results resolve to one of four grades. The grade attaches to the fee, not to the company; a processor's overall scorecard is the aggregate of its graded fees.
A real, external cost passed to the merchant at or near what it costs the processor. Disclosed, honestly named, verifiable. Nothing to answer for.
Processor revenue — and openly so. The service is real, the name says what it is, the price was disclosed, and it has not crept. Profitable is not the same as unfair.
There is a genuine service underneath, but the price, the disclosure, or the pricing structure does not hold up. Typically a fee that has outgrown its justification.
No identifiable underlying cost, a name or description that asserts something untrue, or a charge added without meaningful consent. Reserved for fees that fail Test 4.
How results map to grades. A failure on Test 4 — honest naming — is disqualifying on its own and produces a Junk grade regardless of the other five, because a merchant who has been told the wrong thing cannot evaluate any of the rest. Failures on Tests 1, 2, 3, 5 or 6 accumulate: one failure with a genuine service underneath is Padded; a clean sheet with real processor margin is Fair; a clean sheet on a true external cost is Pass-Through.
The method is fixed, and the order of evidence matters more than anything else in this document. A finding is only as good as where it came from.
We will not publish a merchant's statement without written permission. We will not publish a figure we cannot source. We will not grade a fee we have not read the processor's own description of. If those constraints mean a page sits unfinished, it sits unfinished.
Every processor named in the registry may respond to any grade. Replies are published on the graded page itself, not on a separate page nobody reads, and are not edited for content.
If a processor demonstrates that a finding is wrong, the grade changes and the page carries a dated correction notice permanently. If a processor changes the underlying practice, the grade changes and the page records both the old grade and the date it was revised. A page that has been corrected is more trustworthy than one that never has — the revision history stays visible for that reason.
Grades are reviewed at least annually, and immediately on notice of a change in pricing or disclosure.
The Alliance has two kinds of members, with different obligations.
A merchant signatory states that they expect the six tests to be met by whoever processes their payments, and agrees to be counted. There is no cost, no commitment, and no obligation to change providers or to speak to anyone.
The count is the point. A published grade that a processor can ignore is a blog post. A published grade backed by a named body of merchants is a number that appears in their next earnings call question.
An advisor signatory accepts binding, checkable obligations:
The list is short on purpose, and its length is stated plainly on the page. A directory padded to look established would defeat the only thing it is for.
A standards body that is quietly a sales channel is worse than no standards body, because it spends the credibility of the idea. So the conflicts are stated here, at full size, and repeated on every graded page:
| Founder | Not named at v0. Also owns a for-profit payments advisory. |
| Founder's advisory | Is an advisor signatory, held to §6 like any other, and can be removed like any other. |
| Referral revenue | None. FPA takes no payment from any advisor for listing or referral. |
| Processor funding | None accepted, in any form, including sponsorship, advertising, or event fees. |
| Tax status | None claimed. FPA is an independent project and is not incorporated as a nonprofit. It will say so on the day that changes. |
Two commitments follow from this and are worth stating as obligations rather than intentions. FPA will publish favorable findings about companies the founder's advisory does not work with, including direct competitors, whenever the evidence supports it. And the criteria in §6 were published before the signatory list existed, so that they can be checked against, rather than written around.