payment-choice

---
name: payment-choice
description: Turns a stack of card payment provider quotes, your current statement, and your actual transaction mix into a total cost of acceptance for each option, split into the part that is fixed by law, the part set by the card schemes, and the only part a provider actually competes on, plus a written request for the itemised pricing breakdown the law entitles you to and a like-for-like comparison table. Use when choosing a new card payment provider, when a current one raises its rate, or when a quote arrives that looks cheaper but you cannot tell why.
---

# Choose a card payment provider on total cost, not the rate they quote

You give this your current provider's statement or invoice, any new quotes you have received, and your actual card transaction volume split by debit, credit and contactless where you have it. You get back the total cost of acceptance for each option broken into the interchange fee, which is capped by law and identical everywhere, the scheme fee, which is set by Visa or Mastercard and not something a provider competes on, and the acquirer's own markup, which is the only part a genuine comparison is actually about, plus a written request for the itemised breakdown a provider must give you if you ask, and a comparison table built from that breakdown rather than from headline rates. It will not tell you what a typical effective rate should be for a business like yours, because no published figure exists for a single independent venue and every business's card mix, average transaction value and negotiating position are different.

## What it does

1. **Separate the three components of every card transaction's cost, because only one of them is a provider's own decision.** The interchange fee is paid by the acquirer to the customer's card-issuing bank, and it is capped for UK consumer cards at "0.2% of the value of a transaction for consumer debit cards (including prepaid cards), and 0.3% for consumer credit cards" under the Interchange Fee Regulation, retained as UK law. The scheme fee is charged by Visa or Mastercard for the use of their network, and is set by the scheme, not the acquirer. The remainder, the merchant service charge above interchange and scheme cost, is the acquirer's own margin, and it is the only one of the three a business is actually negotiating when it compares quotes. The Payment Systems Regulator states the point directly for merchants: "the IFR caps certain interchange fees, it does not cap the MSC." A quote that looks cheaper because it undercuts on a component that is fixed by law or set by a scheme is not actually cheaper; it has simply hidden its margin somewhere else in the statement.

2. **Read the current statement line by line, and separate blended from unblended before comparing anything.** A blended rate quotes one percentage across every card type and every transaction, which hides which of the three components in step 1 moved when the rate changed. An unblended, or interchange-plus, statement shows the interchange fee, the scheme fee and the acquirer's markup as three separate lines for each card category. Two providers quoting "1.5%" are not comparable until you know whether that figure includes a debit card interchange of 0.2% and a credit card interchange of up to 0.3%, or excludes them, because the underlying mix of debit and credit in your own trade changes what a blended headline rate actually costs you.

3. **Use the legal right to an itemised breakdown rather than accepting whatever the provider chooses to disclose.** Article 9 of the Interchange Fee Regulation requires that "each acquirer shall offer and charge its payee merchant service charges individually specified for different categories and different brands of payment cards" unless the merchant has separately requested a blended charge in writing, and that the acquirer supply "individually specified information on the amount of the merchant service charges, interchange fees and scheme fees applicable with respect to each category and brand of payment cards." Write to the current provider and to every new quote, asking for this breakdown by name. A provider that cannot or will not itemise interchange, scheme fee and its own markup separately has not given a comparable quote, whatever percentage it has offered.

4. **Use the post-transaction settlement information the law also entitles you to, to check the quote against what is actually being charged.** Article 12 of the same Regulation requires the acquirer to supply, after each transaction, "the amount of any charges for the card-based payment transaction, indicating separately the merchant service charge and the amount of the interchange fee." Where this information is available on request but not supplied automatically, ask for it in writing rather than assuming the monthly statement already shows it broken down this way.

5. **Build the total cost of acceptance from the actual transaction mix, not from a single average rate.** Apply each provider's interchange, scheme fee and markup figures to your real split of debit, credit and contactless volume from the last full quarter, not to an assumed even split. A business whose trade is mostly debit card contactless pays close to the 0.2% floor regardless of provider; a business with a heavier credit card mix, common where corporate accounts or higher average spend are involved, pays closer to the 0.3% floor, and the two businesses will see genuinely different totals from the identical quoted markup.

6. **Add every fee that is not the percentage rate, because this is where a cheaper headline rate is often recovered.** List, for each provider: a monthly minimum or account fee, a PCI compliance or non-compliance fee, a terminal rental or purchase cost, a chargeback handling fee, a payout or settlement delay, and any early termination charge on the current contract. A provider quoting the lowest percentage rate with the highest fixed monthly fee can cost more for a low-volume seasonal business than a provider with a slightly higher rate and no fixed fee, and the total cost of acceptance in step 5 is meaningless without these added in.

7. **Check the contract term and the exit cost before treating a lower quote as a decision, not only a comparison.** A rolling monthly contract has a different total cost of switching than a three-year fixed term with an early termination charge. Where the current provider's contract has time left to run, the exit cost is part of the comparison, not a separate question asked after the decision is made.

8. **Hand the assembly and the comparison table to Codex once every quote and statement is in hand.** Put the current statement, every new quote, the transaction mix, and the answers received to the itemised breakdown request in one working folder and run, from that folder:

   ```
   codex exec --sandbox danger-full-access --skip-git-repo-check -C "<working folder>" - < CODEX.md
   ```

   Claude hands over each provider's interchange, scheme fee, markup and fixed fee figures, and the actual transaction mix. Codex returns the total cost of acceptance for each provider applied to that mix, the comparison table, the itemised breakdown request letter, and a gaps file for any provider that has not supplied a genuine itemised figure. Codex never recommends a provider; it totals what each one has actually disclosed.

9. **Re-run the comparison when the transaction mix changes materially, not only when a contract is up for renewal.** A shift toward more online or phone bookings, a change in average spend, or a new corporate account paying by credit card all change which provider's pricing structure is actually cheapest, because the calculation in step 5 depends on the mix, not only on the quoted rates.

## Then it checks

1. Every provider's figures are split into interchange, scheme fee and acquirer markup, and no provider's total rests on a single blended percentage taken at face value.
2. A written request for the Article 9 itemised breakdown has been sent to every provider under comparison, including the current one.
3. The total cost of acceptance for each provider is calculated against the business's own actual transaction mix, not an assumed or industry-average mix.
4. Every fixed fee, monthly minimum, terminal cost and chargeback fee is listed for each provider and included in the total, not compared on percentage rate alone.
5. The exit cost of the current contract, including any early termination charge and remaining term, is stated alongside the comparison.
6. No file states a typical or average effective rate for a UK hospitality business as a benchmark for what any of the quotes should be.

Any check fails: name it, redo that step once. Failed twice: say what is wrong and stop.

## Rules
- Public information only.
- Never invent a fact, a number or a quote.
- Anything sent in someone's name says whose name it is.
- Never state a typical, average or benchmark effective card processing rate for a single independent UK hospitality business. No published figure exists, and a borrowed rate would be compared against a transaction mix that is not this business's own.
- Never treat a headline blended rate as comparable to another provider's headline blended rate without first breaking both into interchange, scheme fee and markup. A comparison of two blended headline figures alone is not a comparison of cost.
- Never assume a provider's disclosure is complete because a percentage was quoted. Only a figure itemised by category and brand, consistent with Article 9, is treated as a genuine markup figure; anything else is recorded as undisclosed in the gaps file.
- Never recommend a specific named provider. The skill totals the cost of the options supplied; choosing between them, including weighing service quality, support and reliability against cost, is the owner's decision.
- This output is a working document prepared for the owner to check against their own statements and quotes, and for their accountant to check before switching provider or signing a new contract. It applies the published Interchange Fee Regulation and PSR guidance to figures supplied. It is not financial advice and it is not a determination that any quote is accurate or complete.

## Built from
- Regulation (EU) 2015/751 on interchange fees for card-based payment transactions, retained as UK law, https://www.legislation.gov.uk/eur/2015/751/article/9, read 23 September 2026: the unblending requirement and the right to an itemised merchant service charge breakdown, quoted in step 3.
- Regulation (EU) 2015/751, Article 12, https://www.legislation.gov.uk/eur/2015/751/article/12, read 23 September 2026: the post-transaction disclosure of the merchant service charge and interchange fee separately, in step 4.
- Payment Systems Regulator, "The IFR and merchants", https://www.psr.org.uk/our-work/card-payments/the-ifr-and-merchants/, read 23 September 2026: the interchange fee caps and the statement that "the IFR caps certain interchange fees, it does not cap the MSC," in step 1.
- Payment Systems Regulator, "The IFR - what, why and how", https://www.psr.org.uk/our-work/card-payments/the-ifr/, read 23 September 2026: the confirmation that the 0.2% and 0.3% caps apply where the merchant, acquirer and card issuer are all within the UK, in step 1 and step 5.

# payment-choice

## You are given
The current provider's most recent statement or invoice, copied line by line where it breaks out interchange, scheme fee and markup, or noted as blended where it does not. Every new quote under comparison, with whatever pricing detail each provider has supplied, marked by the owner as itemised (interchange, scheme fee and markup shown separately per card category) or blended (a single rate). The business's actual transaction mix for the most recent full quarter, split by debit, credit and contactless where available, in transaction count and value. Every fixed fee each provider charges: monthly minimum, PCI compliance or non-compliance fee, terminal rental or purchase, chargeback handling fee, and any other named fee. The current contract's remaining term and any early termination charge, where one applies. Whether a written itemised breakdown request under Article 9 has been sent to each provider and what response, if any, was received. All judgement calls, including which provider to choose, are left to the owner; you calculate totals only.

## Produce
Write into a `./payment-choice-output/` folder:

1. `provider-breakdown.csv` with columns: `provider_name`, `card_category`, `interchange_rate_percent`, `scheme_fee_percent`, `markup_percent`, `blended_rate_percent`, `disclosure_type`. `card_category` is exactly one of `consumer debit`, `consumer credit`, `commercial card`, `not specified`. `disclosure_type` is exactly one of `itemised per Article 9`, `blended`, `undisclosed`. Where a provider's figures are blended, `interchange_rate_percent`, `scheme_fee_percent` and `markup_percent` are left empty and only `blended_rate_percent` is populated. Never split a blended rate into its components yourself; that is estimation, not calculation.
2. `total-cost-of-acceptance.csv` with columns: `provider_name`, `quarterly_transaction_value_gbp`, `percentage_cost_gbp`, `monthly_fixed_fees_gbp`, `quarterly_fixed_fees_gbp`, `terminal_cost_gbp`, `total_quarterly_cost_gbp`, `effective_rate_percent`. `percentage_cost_gbp` is calculated by applying each provider's rate figures from `provider-breakdown.csv` to the supplied transaction mix, card category by card category where itemised, or the single blended rate where not. `effective_rate_percent` is `total_quarterly_cost_gbp` divided by `quarterly_transaction_value_gbp`, shown as a percentage, and is a calculated result, never a benchmark or a target.
3. `exit-cost.csv` with columns: `provider_name`, `current_provider_yes_no`, `remaining_contract_months`, `early_termination_fee_gbp`, `notes`. One row per provider under comparison; a new quote with no existing contract records `remaining_contract_months` and `early_termination_fee_gbp` as `0`.
4. `itemised-request-letter.md` a request, addressed to each provider by name, asking for the Article 9 itemised merchant service charge breakdown by card category and brand, and the Article 12 post-transaction interchange and scheme fee information, quoting both articles by number. One version per provider that has not yet supplied an itemised breakdown.
5. `comparison-table.md` a single table listing every provider, its `disclosure_type`, its `total_quarterly_cost_gbp` and `effective_rate_percent` from `total-cost-of-acceptance.csv`, and its exit cost from `exit-cost.csv`, sorted by `total_quarterly_cost_gbp` ascending. No recommendation or ranking commentary is added beneath the table.
6. `gaps.md` a numbered list of: every provider with `disclosure_type` of `blended` or `undisclosed`; every card category with no transaction mix data supplied; every provider missing a fixed fee figure; and every provider with no itemised request letter response recorded.

## Rules
- Codex calculates totals from figures supplied. It never estimates a provider's interchange, scheme fee or markup split where only a blended rate was given, never recommends a provider, and never states a typical or average effective rate.
- Never convert a blended rate into estimated components. A blended figure stays blended in every file, flagged as such.
- Never rank or recommend in `comparison-table.md` beyond the sort order stated. The owner decides; Codex totals.
- Use British English, GBP written as `GBP` or the pound sign, and DD Month YYYY dates. No em dashes in any file you write, and any supplied text containing one is recorded verbatim and flagged in `gaps.md`.
- Every provider name, rate and fee is copied exactly as supplied.
- Every file ends with this line: this is a working document prepared for the owner to check against their own statements and quotes, and for their accountant to check before switching provider or signing a new contract. It applies figures supplied by the owner to a calculated comparison and is not financial advice or a determination that any quote is accurate or complete.

## Return
The absolute path of every file written, the row count of each CSV, the total quarterly cost and effective rate calculated for every provider stated plainly with no ranking commentary, the number of providers with `disclosure_type` of `blended` or `undisclosed`, the exit cost of the current provider's remaining contract, and the `gaps.md` item count.