How do payment processors make money? A revenue model breakdown
How payment processors make money: five business models ranked by margin, risk, and the volume each one needs to work.
You see 2.9% + $0.30 on your statement. Your processor keeps a fraction of that.
Of the gross merchant discount rate (MDR) you pay, the largest share is interchange, and that flows straight to the bank that issued your customer's card. Another slice goes to Visa or Mastercard as scheme fees. What's left is the processor's actual margin: often a fraction of one percent on a transaction that looked like 2.9%.
The retained margin is easy to miss, since it's not broken out anywhere you'd normally look. Here's where the money actually goes, and why it changes how you negotiate, route, and choose providers.
What is a payment processor?
A payment processor is the technical intermediary that moves a transaction from the moment a customer clicks โPayโ to the moment funds reach the merchant's account. It handles authorisation, clearing, and settlement โ communicating between the merchant, the card network, the issuing bank, and the acquiring bank in a matter of seconds.
The term is used loosely. In practice, โpayment processorโ describes anything from a pure technical layer earning ~$0.05โ$0.10 per transaction to a full-stack platform that underwrites merchants, manages fraud, and settles funds directly. The business model and the economics differ significantly depending on where in the stack a company actually sits.
What all payment processors share: they donโt own the money moving through them. They facilitate the transfer and earn a fee for doing so. Understanding how that fee is structured โ and what portion the processor actually keeps โ is the starting point for managing payment costs effectively.
5 payment processor business models
Not every company called a โpayment processorโ operates the same way. The label spans five distinct models, each with different margins, risk exposure, and capital requirements. In my experience, where you sit in the stack matters more than any single fee negotiation.
ISV/Referral
The lightest model. You refer merchants to a processor and earn a residual โ typically 5โ15 basis points of processed volume. No card-network registration, no underwriting risk, near-zero fixed cost. The trade-off: you own nothing and earn a fraction of the economics.
Retail ISO
One step up. You resell processing under a sponsor bank's umbrella and earn a residual split, commonly around 30% of the acquirer's markup. The sponsor bank holds the underwriting risk. Fixed costs are low, and so is the ceiling.
PayFac-as-a-Service
Where the economics start to shift materially. You aggregate sub-merchants under your own brand, earning 60โ100 basis points of volume. The sponsor handles PCI DSS compliance, reserves, and sponsorship โ you own the merchant experience and onboarding. This is the most accessible high-margin entry point for payment businesses at $10M+ GPV.
Registered PayFac
Full-stack ownership: your own master merchant account, direct underwriting, and 100% chargeback liability. Gross margin reaches 80โ95 basis points, but fixed costs run $100kโ$750k per year, with 12โ24 months to launch. The economics only work above roughly $500M in annual GPV.
Direct Acquirer/Principal Member
The end state: a principal Visa/Mastercard membership, your own BIN, and direct settlement. You capture the full markup and eliminate the sponsor margin. In return, you take on โฌ50kโโฌ100k in year-one scheme fees, a Payment Institution licence, processing infrastructure, reserve capital, and ongoing compliance. This is infrastructure ownership, not just processing.
The practical implication: most businesses entering payments should start at PayFac-as-a-Service and grow into greater ownership as volume justifies the cost. Jumping straight to a registered PayFac or direct acquiring before hitting the volume threshold means paying for infrastructure that doesnโt yet earn its keep.
Where the money actually goes: revenue line by line
The gross MDR a processor charges looks like revenue. Most of it isn't. In my experience, this is the part most Payment Managers skip โ and it's where the real money is lost. Here is what actually happens to the fee, and who keeps what.
MDR markup/acquirer spread
The core revenue line. After interchange goes to the issuing bank and scheme fees go to Visa or Mastercard, the processor keeps the remainder โ the โplusโ in interchange-plus pricing. For most acquirers and payment platforms, this retained spread is the single largest profit driver. According to industry surveys, mid-tier acquirers earn a median of 64 basis points net per transaction โ with the top 20% earning 103โ128 bps and the bottom 20% just 22โ32 bps.
At the enterprise end, rates tell a different story: the largest global platforms retain as little as 16 basis points of processed volume, because high-volume merchants negotiate hard.
Cutting MDR by 20 basis points feels like a win. It rarely is. Interchange downgrades and scheme fees routinely dwarf the savings โ and at enterprise scale, processors price processing thin on purpose. VAS is where they make the margin back.
Interchange
The biggest line in any MDR flows straight to the card issuer. For the processor, itโs a cost, not revenue, unless they charge a blended or flat rate above actual interchange cost and keep the difference. On a debit transaction at 2.6% flat, that works well (~2.5% margin). On a premium rewards card at the same rate, interchange can exceed 2.0%, leaving the processor with ~0.6% or less.
Under interchange-plus pricing, that card-mix risk shifts to the merchant: they pay exact interchange plus a fixed markup, and the processor's margin is insulated. This is why interchange-plus is preferred by sophisticated merchants โ and why tiered pricing, where the processor routes transactions to the most expensive qualifying category, is widely criticised as opaque and tends to maximise processor margin at the merchant's expense.
FX and cross-border spread
Often the highest-margin transactional line. DCC typically runs 3โ7% above the interbank rate. Cross-border FX spread on settlement adds further. Processors heavily exposed to emerging-market cross-border flows can run net take rates above 1% of volume โ multiples of what a domestic card acquirer retains.
In my experience, this is where the cost of a poorly structured provider relationship shows up most visibly for cross-border merchants โ not in the MDR. The FX line is rarely audited and almost never negotiated in the first conversation.
Value-added services
Where margin compounds at scale. Fraud tooling, 3DS, tokenisation, multi-currency settlement, reporting โ these carry software-like margins and create switching costs. VAS already drives 15โ30% of leading PSPs' net revenue, and the most sophisticated platforms actively grow this share as per-transaction margins compress.
A merchant deeply embedded in a processor's fraud and tokenisation suite is far harder to move than one using raw processing alone, which is precisely why processors invest in it. The stickiness is by design.
Float income
Pure margin โ when it exists. Processors earn interest on merchant and customer funds held between T+1 and T+3 settlement. At scale and in high-rate environments, this is a meaningful revenue line. The catch: float shrinks as settlement accelerates. Any processor marketing instant payouts is directly compressing it, a structural tension between product positioning and margin.
Chargeback fees, gateway fees, and subscription lines
Small individually but compound at volume. Chargeback fees ($15โ$100 per dispute) recover cost and deter abuse. Gateway and monthly fees provide recurring, high-margin software revenue. These lines are routinely under-negotiated by merchants focused solely on MDR, which is exactly why processors rarely highlight them.
Revenue stream | MDR markup/acquirer spread | FX/DCC/ cross-border spread | Value-added services (fraud, 3DS, tokenisation) | Float/settlement income | Chargeback fees | Gateway/ SaaS/subscription | Financing/BNPL |
Margin profile | 30โ100 bps net; varies by stack position | Very high; 3โ7% above interbank | High; 15โ30% of leading PSP net revenue | Pure margin; rate- and speed-dependent | $15โ$100 per dispute | High margin, recurring, sticky | High yield; carries credit risk |
Who captures it | Acquirer, PayFac, ISO | PSP with settlement control | Payment platform/PSP | Balance-sheet players | Processor/acquirer | Payment platforms, ISVs | BNPL providers |
Accessible at launch? | Yes โ the core revenue line | Partially โ requires settlement control | Yes โ highest-margin early expansion | Hard โ needs scale + e-money licence | Yes โ cost recovery + deterrent | Yes โ strong for vertical/ISV models | No โ needs balance sheet or funding partner |
Payment processor pricing models comparison
The pricing model a processor chooses determines how volatile and defensible its margins are. Three models dominate.
- Interchange-plus: exact interchange and scheme fees passed at cost, plus a fixed markup. Processor margin is transparent and insulated from card-mix shifts โ when interchange rises on premium cards, the merchant pays it. Lower raw margin for the processor, but durable and defensible. The preferred model for enterprise merchants who can benchmark markups.
- Flat-rate: the processor keeps the difference between the flat rate and the actual cost. This wins big on debit and low-interchange transactions: collecting 2.6% and paying roughly 0.05% leaves around a 2.5% margin. It loses on premium and commercial cards, where interchange alone can exceed 2.0%, leaving the processor with 0.6% or less. Flat-rate is most profitable when the merchant's card mix skews cheap, and it carries real margin-compression risk if that mix shifts.
- Tiered pricing: categories constructed by the processor, which routes transactions to the most expensive qualifying tier. Produces the highest raw processor margin. Widely criticised as the least transparent model, it tends to generate merchant churn once the markup is discovered.
The highest net revenue outcome combines a defensible processing model with extensive value-added services. Enterprise accounts negotiated close to cost still generate meaningful margin through fraud tooling, tokenisation, multi-currency, and reporting. That is why large-merchant acquisition is economically rational even at thin per-transaction rates.
Model | Transparency for merchant | Processor margin level | Card-mix risk | Best merchant segment |
|---|---|---|---|---|
Interchange-plus | High โ exact costs visible | Lower, but durable | None โ passes to the merchant | Enterprise, high-ticket, B2B |
Flat-rate | Medium โ one rate, no breakdown | Buffer-dependent; compresses on premium cards | High โ processor absorbs premium card cost | SMB, low-ticket, debit-heavy |
Tiered | Low โ tier logic undisclosed | Highest raw margin | Managed via tier routing | Legacy books only |
The blind spots that cost merchants most
Interchange downgrades
Each transaction must meet specific data and timing criteria to qualify for its target interchange rate. Missing AVS/CVV data, settling late, or failing to submit Level 2/3 data on commercial cards triggers a downgrade to a costlier category. Mastercard alone publishes hundreds of interchange rate categories. Most merchants donโt track which category their transactions actually land in โ which means they are paying rates they didnโt agree to and cannot see.
Scheme fees rising faster than volume
The UK Payment Systems Regulator's final report (MR22/1.10, 2025) found that Mastercard and Visa increased their core scheme and processing fees to acquirers by at least 25% since 2017, costing businesses at least ยฃ170 million extra per year. Many scheme fees are non-transactional, billed monthly with a one- to two-month lag, and easy to miss in a per-transaction model. A structured scheme-fee review typically finds a 7โ15% cost-out opportunity.
Chargeback thresholds tightening
Visa's VAMP programme (enforcement from 1 October 2025) charges acquirers $4โ$8 per dispute above thresholds of 0.5% (Above Standard) and 0.7% (Excessive) โ costs passed directly to merchants. Mastercard's ECM/HECM programme imposes fines and a MATCH listing. PayFacs bear 100% of sub-merchant chargeback loss if the merchant cannot pay, making high-risk merchant underwriting a direct P&L risk, not just a compliance question.
Card-mix surprises on flat-rate books
A processing model built on debit assumptions takes a direct margin hit when rewards cards, commercial cards, or cross-border transactions over-index. Under flat-rate pricing, that hit lands on the processor first โ but it eventually surfaces in higher rates, repricing conversations, or unwanted attention to chargeback ratios.
For Payment Managers running high-risk or high-volume operations, the implication is that processor relationship management doesnโt end at the rate you signed. The real management task is monitoring actual economics continuously โ approval rates by provider and card type, decline reasons, scheme-fee growth, chargeback ratios, FX costs โ and having the data to act on what you find.
What this means for your negotiation
Here's what I'd focus on if I were sitting across the table from a processor.
- Fix the leaks before negotiating the rate. The processor's retained margin is a fraction of the headline MDR. Negotiating MDR down by 10 basis points is often less impactful than fixing interchange downgrades, scheme-fee overcharges, or routing gaps that lose 15โ20% of transactions to preventable declines.
- Use approval rates as a commercial argument. A processor's per-transaction margin depends on volume. If your routing generates unnecessary declines, fixing it increases their revenue too. They won't fix it proactively unless you surface the data and frame it as a commercial opportunity. See how smart routing and cascading work in practice.
- VAS attach is leveraged in both directions. If you use a processor's fraud tooling, tokenisation vault, and reporting suite at scale, that attachment is margin for them โ and negotiating room for you. Bundled VAS is where the real pricing conversation happens at enterprise volume, not the MDR.
- Audit your FX costs. For merchants processing across geographies, FX spread and settlement terms often cost more than the MDR, and are rarely reviewed. The payment routing strategy that addresses geo-restrictions, card-issuer behaviour by region, and provider selection for cross-border flows is often the single highest-ROI intervention available.
- High-risk merchants have more leverage than they think. VAS attach rates are higher on high-risk verticals โ fraud tooling, 3DS, risk reporting โ which means there is margin elsewhere in the relationship beyond the MDR. A merchant generating consistent volume and holding chargebacks within scheme thresholds is commercially valuable. Use that.
Want to build on top of these economics? There's a faster way
The traditional path to launching a payment business is expensive and slow. Building processing infrastructure from scratch takes 12โ24 months and millions in capital before you process your first transaction. Licensing, PCI DSS certification, provider integrations, routing logic, fraud tooling, merchant management โ each one is a multi-month project on its own.
The alternative is a white-label payment platform. Instead of building the infrastructure, you license it and launch under your own brand in weeks.
This is what Corefy's white-label solution is built for: PSPs, fintechs, and payment companies that want to enter or scale a payment business without the overhead of building it from scratch.
In practice, that means:
- Your brand, from day one. Your merchants see your logo, your domain, your checkout. The infrastructure runs on Corefy; the business relationship is entirely yours.
- 600+ provider integrations, ready to use. Every acquirer, PSP, and alternative payment method connection is built and maintained by Corefy's team. Adding a new provider takes several clicks.
- Smart routing and cascading out of the box. The same routing logic that lifts approval rates by up to 30% and reduces processing fees โ available from launch, without building it yourself.
- Lower cost than an in-house build. No servers to manage, no six-figure infrastructure spend before you've signed your first merchant.
Letโs talk about how we can help your payment business succeed!
Connect providers, configure routing, and start processing under your brand โ with full infrastructure support from a dedicated payment team.