Interchange is set by the card networks and costs every processor exactly the same. What differs is the markup on top and whether you can see it. At $8M in annual card volume, moving from flat-rate to interchange-plus pricing is worth roughly $54,000 a year — but only if your processor routes transactions cleanly, and your card mix moves the number more than your negotiated rate does.
Most restaurant groups discover their processing cost the way they discover a roof leak: all at once, and later than they would have liked. The monthly statement says 2.9%. The finance lead asks why the same statement said 2.7% in March. Nobody can answer, because a blended rate is an average of things that were never explained.
What is interchange, and who actually sets it?
Interchange is the fee the card-issuing bank keeps on every transaction. Visa and Mastercard publish the rates, the networks set them, and every processor pays the identical amount. No processor has a better deal on interchange, because interchange is not something processors sell.
On top of interchange sit two other layers: network assessments, which are small and also fixed, and the processor's own markup, which is the only part anyone is actually negotiating. When a sales rep quotes you "2.6%", that number contains all three and separates none of them.
If what you need is the definitions, our primer on blended versus interchange-plus pricing covers them. This piece assumes you have them and asks a narrower question: at group volume, what is the difference actually worth?
What are the three pricing models?
| Model | How it is quoted | What you can see | Who it suits |
|---|---|---|---|
| Flat rate | One rate for everything, e.g. 2.6% + 10¢ | Nothing — interchange and markup are fused | Low-volume sites that value simplicity over cost |
| Tiered / blended | Qualified, mid-qualified, non-qualified buckets | Which bucket, but not why a card landed there | Nobody, at scale — the buckets are the processor's own invention |
| Interchange-plus | Interchange at cost, plus a stated markup, e.g. +0.25% + 8¢ | Every layer, itemised per transaction | Groups with the volume to make the markup worth negotiating |
The reason tiered pricing survives is that the tiers are defined by the processor, not the networks. A card can be reclassified into a more expensive bucket without anything about your business changing.
What does $8M in card volume actually cost?
Here is the arithmetic, worked end to end. These are illustrative inputs, not measured results — the point is the shape of the gap and the method, so substitute your own volume, average ticket and card mix and the model still runs.
Assumptions: $8,000,000 in annual card volume, a $42 average ticket (so roughly 190,500 transactions), and a card mix of 40% regulated debit, 35% standard credit, 25% premium rewards credit.
| Card type | Share of volume | Typical interchange | Annual interchange |
|---|---|---|---|
| Regulated debit | 40% — $3.2M | 0.05% + 22¢ | $18,400 |
| Standard credit | 35% — $2.8M | 1.80% + 10¢ | $57,100 |
| Premium rewards credit | 25% — $2.0M | 2.30% + 10¢ | $50,800 |
| Interchange subtotal | $8.0M | — | $126,200 |
| Network assessments | — | ~0.14% | $11,200 |
| Unavoidable network cost | — | 1.72% effective | $137,400 |
That $137,400 is the floor. No processor, and no amount of negotiating, gets you below it. What sits above it is the only thing in play:
| Flat rate (2.6% + 10¢) | Interchange-plus (+0.25% + 8¢) | |
|---|---|---|
| Network cost | $137,400 | $137,400 |
| Processor take | $89,700 | $35,200 |
| Total annual cost | $227,000 | $172,700 |
| Effective rate | 2.84% | 2.16% |
| Difference | — | about $54,000 a year |
Two things are worth noticing. The first is that the processor is earning $89,700 on the flat-rate deal for work that costs it the same either way. The second is that 0.68 of a percentage point does not sound like a negotiation worth having until it is expressed in dollars.
When does interchange-plus cost more, not less?
That $54,000 is what the pricing model is worth if the transactions are submitted cleanly. That is an assumption, not a guarantee, and it is the part most interchange-plus sales conversations leave out.
Every card transaction carries data the network uses to decide which interchange category it belongs in. When that data is incomplete or late — an address check that never ran, a batch settled a day after the sale, a card-present sale submitted as though it were keyed — the network reprices it into a more expensive category. The industry calls these downgrades, and they are common enough to matter.
Which pricing model you are on decides who pays for them:
| When a transaction downgrades | On a flat rate | On interchange-plus |
|---|---|---|
| Who absorbs the higher interchange | The processor — your rate is fixed | You — interchange passes through at cost |
| Whose incentive is it to prevent | The processor's; it comes out of their margin | Nobody's by default — it stopped being their cost |
| Where you would see it | Nowhere. The quoted rate never moves | Your effective rate, with no contract change |
Routing transactions optimally takes real investment in technology, and a processor whose customers cannot tell the difference has little reason to make it. A poor downgrade rate can erase a meaningful share of the saving modelled above — which is the honest reason a well-run blended rate sometimes beats a badly-run interchange-plus deal. The pricing model is half the story; execution is the other half.
There is one more case for staying on a blended rate, and it has nothing to do with arithmetic: if nobody at the group is going to read the statement each month, interchange-plus hands you more information and no one to act on it. Transparency is worth exactly what someone does with it.
Why does your card mix change the answer more than your rate does?
This is the part that catches finance teams out. Regulated debit carries interchange of roughly 0.05% plus 22¢. On a $42 ticket that is about 24¢, or 0.57%. Under a 2.6% flat rate, the same transaction costs $1.19. The processor keeps the difference — and the more debit your guests use, the more it keeps.
| On a $42 ticket | Regulated debit | Premium rewards credit |
|---|---|---|
| Actual network cost | about 24¢ (0.57%) | about $1.07 (2.54%) |
| Cost under 2.6% + 10¢ flat | $1.19 | $1.19 |
| Processor margin on that sale | 95¢ | 12¢ |
A fast-casual or counter-service concept with heavy debit usage is subsidising a pricing model built for the opposite mix. A fine-dining room running mostly premium rewards cards is much closer to break-even on flat rate. Same group, same contract, two very different outcomes — which is why group-level averages hide the problem rather than reveal it.
The regulated debit cap comes from the Durbin Amendment and applies to issuers with more than $10 billion in assets. Cards from smaller banks and credit unions are exempt and price higher, so the debit share of your mix is not uniformly cheap either.
How do you read a merchant statement?
Most of the cost information a group needs is already arriving every month; it is just arranged to discourage reading. Four numbers do almost all the work.
| Find this | Where it hides | What it tells you |
|---|---|---|
| Total card volume processed | Usually page one, near the deposit summary | The denominator for everything else |
| Total fees charged | Often split across processing, service and "other" sections — add them all | The numerator, including the fees nobody quoted you |
| Effective rate | Almost never printed — divide total fees by total volume | The only rate worth comparing between processors |
| Interchange detail | An appendix, sometimes only on request | Whether you are on interchange-plus at all |
If the statement cannot tell you what share of your fees was interchange and what share was markup, that is itself the finding. Run the effective-rate calculation for each location separately before you average it: a group number conceals exactly the variation that makes the case for renegotiating.
Do this for three consecutive months. A rate that drifts while your contract sits still is card mix moving, not a billing error — and it tells you which locations are most exposed.
What should a group ask a processor?
| Question | What a good answer sounds like | What should worry you |
|---|---|---|
| Can I see interchange and your markup as separate line items? | Yes, on every transaction, in the statement itself | "We can pull that for you on request" |
| What is the markup, stated as basis points plus cents? | A single number that does not change by card type | A range, or a rate that varies by "qualification" |
| What is my downgrade rate, and will you report it monthly? | A number, how it is measured, and a standing line item | "Downgrades aren't really a thing any more" |
| What is the effective rate across my whole portfolio? | Total cost divided by total volume, per location | An average quoted without the denominator |
| Which fees are not in the rate? | A complete list: monthly, PCI, gateway, batch, chargeback | "That's it" |
| What happens at renewal? | A stated term and a stated notice period | Auto-renewal with a termination fee |
| Can I keep this processor if I change POS? | Yes | A shrug |
That last question is the one groups forget to ask, and it is the one that determines whether any of the others matter in two years. We covered the wider version of this in how to move a restaurant group to a new POS without closing a location.
Where does the POS fit into this?
It decides whether you have a choice at all. A POS that bundles payment processing and will not run on anything else has effectively set your rate for you, and the pricing conversation is over before it starts. Interchange-plus is only worth negotiating if you are free to take the negotiation elsewhere.
GoTab is processor-agnostic: a group can bring its own processor and its own negotiated rate, and keep them through a POS change. Software is priced per device rather than as a share of what you process, so the POS cost and the payments cost stay separable — which also means the two can be renegotiated independently instead of moving as one block.
For where this sits in a wider platform decision, see what restaurant groups should look for in an enterprise POS, or the point of sale platform itself.
Restaurant credit card processing, answered briefly
Is interchange-plus always cheaper than a blended rate? No. It is always more transparent, which is a different claim. Because it passes network costs through at cost, it also passes through the cost of badly routed transactions. A tightly priced blended rate from a processor with good routing discipline can beat a loose interchange-plus deal — particularly for a group with nobody assigned to read the statement each month.
What is a good credit card processing rate for a restaurant? The honest answer is that the rate is the wrong unit. Ask for the effective rate — total processing cost divided by total card volume — and compare that. A group on interchange-plus with a tight markup typically lands somewhere near 2.1–2.3% effective; flat-rate deals usually sit around 2.7–2.9%.
At what volume does interchange-plus start to pay off? It scales with the dollars, not the location count. A single high-volume site can save more than five small ones. The practical threshold is less about volume than about whether anyone has the time to read the statement — interchange-plus gives you more information, which is only an advantage if someone uses it.
Is surcharging a better answer than renegotiating? They are different levers and the rules differ by jurisdiction. Surcharging moves the cost to the guest; interchange-plus reduces it. Most groups should exhaust the second before touching the first.
Why did my effective rate go up without anything changing? Usually card mix. A shift toward premium rewards cards raises blended cost with no change to your contract — which is exactly the movement a blended rate is unable to show you.
Do I have to change POS to change processor? Only if your POS requires it. Some platforms bundle the two; others let you keep your processor through a POS change. Establish which you have before you negotiate, because it determines your leverage.
Want to run this arithmetic against your own volume and card mix? Request a demo and we will work through your statement with you.
