Pay by Bank vs card fees: when does A2A actually pay off?
A UK operator's Head of Payments runs the same calculation every month. Volume on plan. Average ticket on plan. Take rate creeping. The third statement in a row to come in heavier than budget, and no operational change to point at. Just the rail, doing what rails do, quietly stacking percentages onto a line item that grows with the business.
The cost question turns on one number most pitches skip past. Above roughly 35 EUR average ticket, Pay by Bank is structurally cheaper than card. Below it, card wins. That is the answer in one sentence, and the rest of this post is what stands behind it.
Our first post laid out the broad case for Pay by Bank across cost, security, and speed. This one stays on cost, because that is the line item that compounds without permission.
TL;DR
- Above ~35 EUR average ticket, A2A is structurally cheaper. Below it, card pricing usually wins.
- Card fees scale with transaction value. Pay by Bank is fixed-fee per transaction.
- Scheme fees have risen sharply for large UK merchants over the last five years.
What is account-to-account (A2A) payment? A2A moves funds directly between a customer's bank account and a merchant's, via Open Banking APIs, without routing through a card network. The customer authenticates in their own banking app. No card details are entered, and no scheme network sits in the middle.
Why the threshold exists
Cards are percentage-based by design. Interchange, scheme fees, and acquirer margin all sit as a share of the transaction value. Bigger ticket, higher cost per transaction. A merchant whose average ticket has climbed steadily over the last five years now pays more per transaction in card fees, for the same product, the same checkout, the same operational footprint. And the scheme fees themselves have risen sharply in that period.
Pay by Bank does not have that mechanic. The fee is fixed per transaction, set independently of the value. A small deposit and a large deposit cost the same to process. At low ticket, the fixed fee is the higher percentage. At high ticket, it is dramatically lower. The two curves cross around 35 EUR, roughly the price of a casual restaurant meal.
What sits above and below the line
iGaming and betting deposits sit well above the threshold. Trading platform funding events run larger again. For these verticals, the cost gap compounds over annual volume.
Convenience retail, food delivery, and micro-transactions sit below. The card percentage on a small ticket is lower than a fixed A2A fee. No reframing changes that.
The honest version of the cost case is dependency, not universality. The math is not in dispute. The threshold is.
The conversion side
The cost case does not stand alone. Returning depositors authenticate through a saved token that cuts the repeat-deposit flow to roughly half the click-through of a first-time payment. For merchants whose revenue depends on repeat volume, the cost saving sits next to a conversion lift on the same rail.
Run your own numbers
Enter your monthly transaction count, your average payment amount, and your current blended card rate. The calculator returns indicative monthly and annual savings. Where the math does not land, it tells you so: "current card pricing may be cheaper at this payment amount."
What the math actually tells you
Back to the operator's Head of Payments. If their average ticket sits above the threshold and their blended card rate has climbed over the past few years, the saving is a P&L line they can defend at the quarterly review, and a way to route around a cost item that was set to keep climbing. If it sits below, the rail is not the lever; the conversation moves elsewhere.
The rail is an AOV decision.
