Interchange-Plus vs Flat-Rate vs Tiered Pricing

Three ways to price the same transaction

Two businesses can run identical transactions on identical hardware through the same underlying processor and pay materially different amounts, not because their costs differ but because their contracts describe those costs differently. The pricing model is not a detail of the deal. It determines what a merchant can see, what it can verify, and what happens to its costs when the card networks change their schedules.

A fourth shape has become common enough to name: subscription or membership pricing, where the merchant pays a fixed monthly fee plus interchange at cost plus a flat per-transaction amount, with no percentage markup at all. Stax Payments built its business on that structure and Finix states that it charges for software and per transaction rather than taking a percentage of volume. Structurally it is interchange-plus with the percentage component moved into a subscription.

Interchange-plus: cost, plus a number you can read

Under interchange-plus, sometimes written as interchange++ where network fees are itemized separately from interchange, the processor passes through the actual interchange and the actual network assessments for each transaction and adds a stated markup — typically a percentage of volume plus a fixed amount per transaction. The statement shows each interchange category, its volume, and its cost, then shows the markup as its own line.

Three properties follow from that, and they are the entire case for the model.

• It is auditable. The interchange lines can be checked against the network's published schedule. Nothing else in payments can be checked against a public document.
• Cost reductions reach the merchant. When a transaction qualifies for a cheaper category — because the batch closed on time, because Level 3 data was passed, because the customer used regulated debit — the saving lands on the merchant's statement. Under any blended model it lands on the processor's.
• Comparison becomes trivial. Two interchange-plus quotes differ by exactly two numbers, the percentage and the per-transaction amount, because the pass-through component is identical for both.

The model has real costs. Statements are long and require someone to read them. The monthly total varies with card mix, which finance teams dislike. And it is only meaningfully available to merchants whose volume is worth quoting individually — though that threshold has fallen sharply, because providers such as Helcim now publish an interchange-plus margin table openly with volume discounts applied automatically, and Adyen passes interchange and scheme fees through at cost with its own processing fee stated separately.

Flat-rate: one number, and a pooled average

Under flat-rate pricing, the merchant pays a single percentage plus a fixed amount per transaction, usually differentiated only by channel — tapped or dipped in person, online, keyed by hand, sent by invoice. The processor absorbs the variance between what different cards actually cost and what the merchant pays. Square, Stripe on standard terms, Shopify Payments, PayPal's standard merchant pricing, and the published small-business rates of several bank acquirers all work this way.

What the merchant is buying is not a low price. It is three other things: predictability, since the cost of a sale is knowable at the moment it happens; simplicity, since there is nothing to audit; and instant onboarding, since these providers are typically payment facilitators that board merchants as sub-merchants under a master account rather than underwriting each one individually.

What the merchant is giving up is the entire upside of its own card mix. A flat rate is priced off a portfolio average. A business whose customers pay largely with regulated debit — grocery, convenience, quick service, discount retail — has a genuinely cheap card mix and is paying the portfolio price for it. A business whose customers pay with premium rewards credit cards is on the other side of that trade and is quietly being subsidized. Neither business can see which one it is, because the flat rate is designed so that it cannot.

Tiered pricing: buckets defined by the processor, not the network

Under tiered pricing — also sold as bundled or qualified pricing — the processor sorts every transaction into a small number of buckets, conventionally qualified, mid-qualified, and non-qualified, and charges a different rate for each. The quoted rate on the proposal is always the qualified rate.

Here is the fact the model depends on the merchant not knowing: those tiers do not exist in the card networks' schedules. Visa and Mastercard publish hundreds of interchange categories; they do not publish a qualified tier. The buckets are invented by the processor, the rules assigning transactions to them are written by the processor, and those rules appear in no public document and generally not in the merchant agreement either. The processor decides both what each bucket costs and which transactions land in it.

The consequences compound.

1. The quoted rate describes a minority of volume. Rewards credit cards, corporate cards, keyed transactions, international cards, and anything with a data imperfection are routinely assigned to the higher tiers. For many merchants the qualified bucket carries well under half of total volume.
2. Downgrades become invisible. A merchant on interchange-plus can see a late batch or a missing Level 3 field as a specific interchange category on the statement and fix it. A merchant on tiered pricing sees only more volume in the non-qualified bucket, with no way to distinguish an avoidable data error from a genuine premium card.
3. Interchange reductions never reach the merchant. When a network lowers a category, or when the merchant's own transaction quality improves, the wholesale cost falls and the tier rates do not. The processor keeps the difference.
4. The bucketing rules can be changed unilaterally. Because they are not disclosed, a processor can reclassify transaction types and increase its own margin without changing a single published rate. There is no line on the statement where this would appear.

Say it plainly: tiered pricing is the wrong choice for almost every merchant, and its defining feature is not that it is expensive but that it is unauditable. A merchant cannot verify a tiered statement against any external reference, which means it cannot detect being overcharged, cannot measure the effect of an operational fix, and cannot compare two tiered quotes against each other. Any provider that will not say what defines a qualified transaction has answered the question.

A worked comparison — illustrative numbers only

The numbers below are hypothetical and for illustration only. They are not current market rates, not quotes, and not representative of any named provider. They exist to show how the three models behave on identical volume, which is the part that does not change even as the real numbers do.

Take a hypothetical merchant processing $100,000 a month across 2,000 transactions, an average ticket of $50. Assume its true wholesale cost — interchange plus assessments, blended across its actual card mix — comes to 1.85% of volume plus $0.11 per transaction. That is $1,850 plus $220, or $2,070 of unavoidable cost. Every model is pricing that same $2,070.

Model A: interchange-plus at a hypothetical 0.30% + $0.10

• Pass-through cost: $2,070
• Markup: 0.30% of $100,000 = $300, plus $0.10 x 2,000 = $200, total $500
• Monthly total: $2,570

Model B: flat-rate at a hypothetical 2.9% + $0.30

• 2.9% of $100,000 = $2,900, plus $0.30 x 2,000 = $600
• Monthly total: $3,500, of which $1,430 is the processor's spread

Model C: tiered, quoted at a hypothetical qualified rate of 1.79%

Assume the processor's undisclosed rules place 55% of volume in qualified, 25% in mid-qualified at 2.55%, and 20% in non-qualified at 3.45%, with $0.20 per transaction and $85 of monthly account, statement, and PCI fees.

• Qualified: $55,000 x 1.79% = $984.50
• Mid-qualified: $25,000 x 2.55% = $637.50
• Non-qualified: $20,000 x 3.45% = $690.00
• Per-transaction: $0.20 x 2,000 = $400.00
• Fixed monthly fees: $85.00
• Monthly total: $2,797

Two lessons survive any change to the inputs. The lowest headline number produced the middle result, which is the entire commercial function of a qualified rate. And under interchange-plus the merchant can see the $500 it is paying the processor, while under the other two models that figure does not appear anywhere and must be inferred.

Effective rate: the only number that compares them

Headline rates from different models are not comparable quantities. The number that is comparable is the effective rate.

To compute it from a statement:

1. Find total sales volume processed for the month — gross card sales, not net deposits.
2. Find total fees charged. On gross funding this is a single monthly debit. On daily net funding the deposits are already net, so the statement's fee summary is the number to use.
3. Divide fees by volume and multiply by 100.
4. Repeat for at least three consecutive months, because card mix and downgrade behavior move seasonally and a single month is not evidence.

Two cautions. First, the effective rate is a diagnostic, not a verdict: a business with a $6 average ticket will always show a high effective rate because per-transaction fees dominate, and a business selling in thousands will always show a low one. Compare a business against its own prior months and against businesses with a similar ticket size, never against a rate quoted in an advertisement. Second, an effective rate calculation is only actionable when the statement itemizes. Knowing you pay 2.80% tells you nothing about which component to attack unless you can see the interchange line separately.

Gravity Payments publishes an effective-rate calculator for exactly this purpose, which is a reasonable indication of how routinely the number is misunderstood by merchants who have been accepting cards for years.

Subscription pricing, and where it makes sense

Subscription or membership pricing removes the percentage markup entirely: a fixed monthly fee, interchange and assessments at cost, and a flat per-transaction amount. On the illustrative merchant above, a hypothetical $99 monthly membership plus $0.08 per transaction would produce $2,070 + $99 + $160 = $2,329, or a 2.33% illustrative effective rate — better than any of the three models compared earlier.

The structure has a threshold built into it. Because the fee is fixed, it must be recovered against the percentage markup a competitor would have charged. Below the crossover volume the subscription is simply worse, and the crossover moves with average ticket: a business with a large average ticket reaches it sooner, because a percentage markup costs more on a $400 sale than a flat per-transaction fee does. Any merchant considering the model should compute its own crossover point rather than accept the provider's, and should recompute it if volume falls — a seasonal business can be on the wrong side of it for half the year.

How to read your own statement

Everything above is only usable if the statement can be interpreted. Work through it in this order.

1. Identify the model. If you see interchange categories listed by name with volumes against them, you are on interchange-plus. If you see qualified, mid-qualified, and non-qualified, you are on tiered. If you see one rate applied to everything, you are on flat-rate. If the statement resists this classification, that is itself the finding.
2. Compute the effective rate for three consecutive months.
3. Split fees into pass-through and markup. Interchange and network assessments in one column, everything the processor keeps in the other. On a tiered statement this split cannot be made, which is the argument for leaving.
4. Convert the fixed fees to a rate. Sum every monthly charge, divide by monthly volume. A modest list of small fees can exceed the entire percentage markup at low volume.
5. Look for volume in expensive categories that should not be there. Late-settlement and missing-data downgrades are operational faults, not pricing faults, and no renegotiation fixes them.
6. Check the contract terms alongside the rate. Term, auto-renewal, early termination amount, and any equipment lease. A lower rate inside a three-year agreement with a non-cancelable terminal lease is not a lower cost.

Which model to be on

A defensible default, stated as plainly as the evidence allows:

• Low or irregular volume, or nobody to read a statement: flat-rate. The spread is real but the certainty and the instant onboarding are worth it, and the absolute amounts are small.
• Steady volume, a debit-heavy or mixed card base, or business-to-business sales: interchange-plus. The pass-through advantage is largest exactly where the card mix is cheapest, and it is the only model in which a merchant can act on what it learns.
• High volume with a large average ticket: compare interchange-plus against subscription pricing on your own three months of data, not on a provider's example.
• Tiered: no. If you are already on it, the migration is worth doing on transparency grounds alone, before any argument about price.

And one habit worth more than the choice itself: recompute the effective rate every quarter and re-quote every year or two. Payment pricing does not decay because anyone acted in bad faith. It decays because a rate negotiated against one year's volume, ticket size, and card mix is still sitting there three years later, and nobody looked.

Frequently asked questions