The Effective Card Rate Is Not the Rate You Were Quoted

Why this matters

Every shop knows the rate it was quoted. Very few know the rate it actually paid, and the gap is not small, not an accident, and not anybody cheating. The realized rate is total card processing fees divided by total card money collected in the period, computed after the fact, and it runs above the headline for four structural reasons that are all predictable and three of which you can size yourself in an afternoon. Until you have computed it, every decision that touches card acceptance - a minimum, a steering policy, a renegotiation, a surcharge - is being made against a number that is not real.

What the realized rate is computed over

Total card processing fees in the window, divided by total card money collected in the window. Both halves are anchored on the payment date, not the invoice date, because a fee is charged when the transaction runs.

Two consequences worth stating before any arithmetic:

• Fees on a refunded payment stay taken. The transaction happened, the processing happened, and most processors do not return the acquirer's fee on a refund. So a month with a large refund shows a realized rate computed on money that partly went back out. Read the fee line and the refund line separately rather than trying to net them.
• A chargeback is not a refund and does not belong in either half. It arrives with its own fee and its own timeline. Keep it out of this calculation or it will look like your processing cost jumped.

Which statement lines go in the numerator

Decide this once and write it down, because changing it later silently breaks every comparison you have.

In: interchange, assessments, the processor's variable percentage, the fixed per-transaction amount, any downgrade or tier surcharge, and any platform points. Everything that scales with the money you took.

Out, and carried as a separate monthly line: the monthly account fee, terminal rental, PCI compliance or non-compliance fees, statement fees, and chargeback fees. These are fixed costs of acceptance, not a rate.

The reason to keep them out is arithmetic rather than tidiness. Suppose your fixed monthly card costs equal 0.25 points on a normal month's card volume. In a month running at 60 percent of that volume they equal 0.25 divided by 0.60, which is 0.42 points. Fold them into the rate and your processing appears to get 0.17 points worse every slow season, forever, for no reason connected to how you take payments. Keep them separate, and add them back once a year to get an all-in cost of acceptance, which is a different and also useful number.

The four reasons it runs above the quote

One: the fixed per-transaction component. Nearly every card price has a variable percentage plus a fixed amount per transaction. The variable part is the part everyone quotes. Money-weighted, the fixed part adds the fixed amount divided by your average card ticket, and per ticket it adds the fixed amount divided by that ticket, which is a completely different number on a small job. The tell: your realized rate is worst in the months with the most transactions and the least money.

Two: card type mix. Rewards cards, small-business cards and corporate cards cost more to accept than a plain consumer debit or credit card. On a flat-rate offer they surface as a higher tier; on interchange-plus they surface as a higher interchange line. The tell: your interchange or tier line moves while your volume does not.

Three: card not present. A payment taken over the phone, keyed by hand, or paid through a link prices above one dipped or tapped at the door, because the card was never physically read. The tell: the premium tracks your link-and-phone share exactly, so splitting card money by channel resolves this one in a single month.

Four: any platform or software component riding on top. Some field-service and payment platforms add their own points to the processing cost. The tell: it is the only one of the four that does not move when your mix moves. It is the same number of points every month regardless of what you sold.

One month's gap, built up line by line

This shop was quoted a flat 2.60 percent plus a fixed per-transaction amount. Its realized rate for the month is built up below. All figures are in percentage points of card money collected; the final row is the total, not another addition.

The gap to the quote is 1.27 points, which is 49 percent above the quoted 2.60. The two largest additions are the fixed component at 0.53 points and the platform component at 0.50 points; between them they are 0.53 plus 0.50, so 1.03 of the 1.27-point gap, which is 81 percent of it. The two mix premiums together contribute 0.24 points, which is 19 percent of the gap.

The small-ticket effect, per ticket and money-weighted

Take 100 card transactions in this month, sized as multiples of the shop's average ticket. The fixed per-transaction component equals 0.5 percent of that average ticket.

The money shares are derived, not assumed: 38 transactions at 0.15x is 5.70 ticket-units, 34 at 0.55x is 18.70, 21 at 1.8x is 37.80, and 7 at 4.5x is 31.50, totalling 93.70 ticket-units across 100 transactions. So the average card ticket is 0.937 times the shop's overall average ticket, and each band's share of card money is its own ticket-units over 93.70.

Now both readings, and they are different units of analysis, which is the whole point.

Per ticket, a job in the bottom band realizes roughly 2.60 plus 3.33 plus the 0.50 platform component, which is about 6.4 percent, before any card-type or channel premium. That is about 1.7 times the shop's blended 3.87 percent, on the same shop, in the same month. Anyone who works out that number for the first time reaches straight for a card minimum.

Money-weighted, those same tickets are 6 percent of card money. Total fixed cost is 100 transactions times 0.005 ticket-units, which is 0.50 ticket-units, over 93.70 of card money: 0.534 percent, which is the 0.53 points in the table above. The 3.33 percent figure is true of a ticket and the 0.53 figure is true of the book, and quoting either one where the other belongs is how this decision goes wrong in both directions.

Three roads, and only one of them is a fee decision

Size the same change three ways before picking one.

Road A: refuse card below a quarter of your average ticket. Those 38 transactions leave, taking 6 percent of card money with them. Fixed cost falls to 62 times 0.005, which is 0.31 ticket-units, over the remaining 88.00: 0.352 percent. The blended fixed component improves by 0.534 minus 0.352, which is 0.18 points, taking the realized rate from 3.87 to 3.69 percent. What it costs: 38 percent of your card transactions now have to pay another way, and those are disproportionately the fast door payments, so you have traded fee points for days to pay and for collection labour on the ones that do not pay at the door.

Road B: take 0.18 points off the variable rate or off the platform component. Identical effect on the blended rate. One conversation, no customer changes anything, and no technician has to have an argument on a doorstep. If the platform component is 0.50 points, this is the first place to look, because it is the one line that is pure price rather than mix.

Road C: raise the smallest tickets rather than refusing them. Put a minimum charge or a trip fee in place that moves the bottom band's midpoint from 0.15x to 0.30x. Fixed cost per ticket in that band halves, from 3.33 percent of the ticket to 1.67 percent. The blended effect is small: those 38 transactions now carry 11.40 ticket-units, total card money rises to 99.40, and the blended fixed component becomes 0.50 over 99.40, which is 0.503 percent - an improvement of only 0.03 points. But the revenue on 38 transactions doubled, and the per-ticket economics that made the bottom band look alarming are gone.

The lesson in the three roads is that the blended rate is the wrong lens for a minimum-charge decision. Road A looks best on the blended rate and is the worst of the three for the business. Road C barely registers on the fee line and is almost always the right answer.

The decisions, and the one that is gated

• A card minimum. Federal law, the Durbin Amendment at 15 U.S.C. 1693o-2(b)(3), lets a US merchant set a minimum of up to 10 dollars on CREDIT card transactions and lets the Federal Reserve raise that figure, which it has not done. The same provision gives you no minimum at all on debit, so a debit minimum is a merchant-agreement breach rather than a smaller version of the same right, and your own agreement can be stricter than the statute. Size it with Road A above rather than by instinct.
• Steering. A discount for paying by cash or check is far less constrained than surcharging, and federal law at 15 U.S.C. 1666f protects it, but only where the discount is offered to every prospective buyer, disclosed clearly, and taken OFF a posted or quoted price that is the card price. Post the cash price and add a fee at the register and you have built a surcharge wearing a discount's name, which inherits every restriction in the bullet below. Say it once, at the point of quoting, not at the point of paying.
• Surcharging is the gated one. It is prohibited outright in some jurisdictions and restricted in others, and the card networks impose their own conditions on top, in whatever edition of their rules is current when you file: typically 30 days advance written notice to your acquirer and to each network, a cap set at your actual cost of acceptance and in no case above the network's own ceiling (the major networks publish ceilings in the 3 to 4 percent range and have moved them), disclosure at your premises, at the point of sale and on the receipt, and credit cards only, never debit or prepaid even when run as credit. Treat it as a question for your acquirer and your own attorney rather than a switch you flip: confirm the current notice period and the current ceiling with them before you file rather than from this card, and do not copy what a shop in another state does.

What changes the answer: a shop whose card money is mostly card present at the door has almost no channel premium to recover, so Roads A and C move nothing worth chasing and the only real lever is Road B. A shop whose card money is mostly linked or keyed should fix the channel before renegotiating anything, because moving payments to the door is usually worth more points than any rate conversation.

References

• Your card network merchant rules and your acquirer's merchant agreement, which set the surcharging conditions above and can be stricter than the statutory credit-card minimum.
• See related: universal-what-share-of-collections-arrives-by-card, for the channel split this card depends on and the arrival-date anchor both figures share.
• See related: universal-days-to-pay-by-method-and-what-the-gap-buys-you, which puts this realized rate against the speed it buys.