Merchant ARPU by Region: How to Compare Payment Markets

Merchant ARPU by region is comparable only when revenue attributable to the same type of merchant cohort is divided by the average number of those merchants over the same period. Before comparing markets, align the numerator, denominator, period, currency, segment and accounting basis. Public disclosures rarely satisfy all six conditions, so many apparent geographic gaps are measurement gaps rather than proven differences in merchant worth.

Merchant ARPU by region is comparable only when the units match

Merchant ARPU is a useful operating measure, but the acronym can hide material differences. For a PSP or acquirer, the cleanest formula is:

Merchant ARPU = revenue attributable to a defined merchant cohort during the period ÷ average active merchants in that cohort during the same period.

The average merchant count should normally be the opening count plus the closing count, divided by two. A monthly average is better when merchant additions, closures or acquisitions are uneven. Using only the closing count can overstate or understate ARPU when the estate changes materially during the year.

The numerator also needs a clear boundary. Total company revenue may include hardware, lending, issuing, tax-free shopping, professional services or businesses that do not share the reported merchant denominator. Segment revenue is preferable when the denominator describes that same segment. Net revenue can be more informative than gross revenue where gross presentation includes network-fee pass-through, but it must not be silently substituted for another issuer's gross revenue.

The following terms must remain separate:

• Revenue is income recognised under the company's accounting policies. Its gross or net presentation depends partly on whether the company acts as principal or agent for particular fees.
• Annual recurring revenue, or ARR, is a run-rate measure of recurring contracted or monetised revenue. It is not the same as revenue recognised during a financial year.
• Gross profit is revenue less the direct costs classified as cost of sales. Gross profit per merchant measures contribution after those costs, not revenue per merchant.
• Merchant may mean a contracted legal entity, an active accepting business or an end-business served through a platform. The issuer's definition governs.
• Outlet or location usually means a physical or billing site. One merchant can operate many locations, while one location can have more than one payment relationship.
• MID, or merchant identifier, is an acquiring or processing account identifier. A legal merchant or outlet may have several MIDs by channel, currency, acquirer or business line.
• Terminal is a device. Multiple terminals may operate at one outlet, so terminal count is not a substitute for merchant or location count.
• Active user or customer can include consumers, bill-payment users or platform end-customers. It is not a merchant denominator unless the source explicitly defines it that way.

This is why Fawry’s disclosed active-network-customer count must not be used to calculate merchant ARPU. Dividing company revenue by that consumer or network-user measure would produce revenue per active network customer, not revenue per merchant. It belongs in a different analysis with a different label.

Geography cannot be isolated until four reporting effects are controlled

A difference between two reported ARPU figures may reflect geography, but it may also reflect four other factors that move at the same time.

First, local commercial conditions matter. Average transaction values, payment-method mix, merchant service charges, local competition, scheme economics and regulation can affect provider monetisation. These factors should be tested with local data rather than inferred from country income alone.

Second, merchant mix matters. A portfolio of enterprise restaurants, small retailers, online sellers and micro-merchants will not produce one stable economic profile. Merchant Category Codes can help segment the portfolio, but an MCC does not reveal outlet count, contract structure or product attachment by itself.

Third, product attachment matters. Payment acceptance, EPOS subscriptions, terminal rental, reporting tools, risk services and other value-added services can contribute different revenue streams. Their presence may explain part of a provider's monetisation, but a cross-company observation does not prove that one product caused higher ARPU.

Fourth, reporting construction matters. Revenue can be presented gross or net, acquisitions can change both numerator and denominator, and issuers can define geographies using billing address rather than the location of the transaction or terminal. Adyen, for example, states that its regional net-revenue figures use the billing location requested by the merchant. Its FY2025 annual report therefore cannot be read as a distribution of terminals, stores or transactions by region. Adyen reported €2,364.2 million of FY2025 net revenue, but did not disclose a total merchant count suitable for a global merchant-ARPU calculation in the reviewed reports. See the Adyen FY2025 Annual Report, year ended 31 December 2025.

A geographic comparison should begin only after these reporting effects have been controlled or clearly marked as unresolved.

Public disclosures belong in separate metric groups, not one ranking

The evidence below shows why apparently useful public figures cannot be placed in a lowest-to-highest merchant-value league table.

Lightspeed is the clearest example because it defines and discloses ARPU directly. Its FY2026 Annual Report, year ended 31 March 2026 reports monthly ARPU of approximately $602 per Customer Location, calculated by the issuer from subscription and transaction-based revenue divided by average Customer Locations. The annualised $7,224 figure above is a calculation, not an issuer-disclosed annual ARPU.

The definition still limits comparison. Lightspeed says a Customer Location is a billing merchant location whose service term has not ended or is under renewal negotiation. For NuORDER, the unit can instead be a brand with a paid subscription. At 30 June 2026, the company separately disclosed approximately 146,000 Customer Locations after the Upserve disposal in its Q1 FY2027 earnings release, quarter ended 30 June 2026. That closing count should not be substituted for FY2026's average denominator.

Shift4 illustrates a different problem. Its FY2025 Form 10-K, year ended 31 December 2025 reports $4.180 billion of gross revenue, $1.981 billion of gross revenue less network fees, a non-GAAP measure, and $1.354 billion of US GAAP gross profit. The filing does not provide a consolidated active-merchant or merchant-location count. Choosing one numerator and attaching a narrower restaurant or acquired-estate count would create a synthetic metric with mismatched scope.

These examples are not a ranking of Lightspeed, Adyen, Shift4 or Fawry. They are a demonstration that revenue, gross profit and ARR must stay in separate metric families, just as merchants, locations, outlets, MIDs and network customers must stay in separate denominator families.

A reproducible comparison needs eight checks before calculation

Use the following checklist for every company, country and reporting period. If a field is unknown, mark it unknown rather than estimating it from a different source population.

1. Define the entity and segment. Record the legal issuer, operating segment and products included in the numerator.
2. Fix the period. Use the same period length and record the fiscal year-end. Do not mix a quarter, annual run rate and audited full year.
3. Name the numerator exactly. State whether it is gross revenue, net revenue, ARR, gross profit or segment revenue. Preserve the issuer's accounting label.
4. Name the denominator exactly. State whether it is average active merchants, closing merchants, outlets, Customer Locations, MIDs, terminals or active users.
5. Align scope and timing. Confirm that numerator and denominator cover the same segment, geography and months. Prefer average counts over closing counts.
6. Keep local currency first. Compare operating trends in reporting currency. If conversion is necessary, publish the FX source, rate date and formula. Do not mix issuer-supplied conversion dates.
7. Label the value's status. Mark each number as issuer disclosed, calculated, estimated or issuer claimed. Show every formula for calculated values.
8. Assign a comparability decision. Green means the same metric definition, scope, period and denominator. Amber means a useful comparison with explicit differences. Red means do not rank, average or infer a gap.

A workbook built from these fields is auditable. It also prevents a common mistake: converting every number to US dollars before discovering that the underlying units were never comparable.

Merchant worth requires gross profit, retention and cost to serve

Revenue per merchant is not merchant lifetime value. A useful merchant-worth model should move from revenue to contribution and then to duration:

Indicative merchant lifetime value = periodic gross profit attributable to the merchant × expected retained periods, adjusted for acquisition, onboarding, hardware, support and servicing costs.

This is an analytical framework, not a universal accounting formula. Contract duration is not the same as realised retention, and a net increase in merchant count cannot reveal gross additions and gross losses separately. Where cohort retention is unavailable, the model should use scenarios rather than a single precise lifetime.

At minimum, a PSP or acquirer should model:

• payment and non-payment revenue by merchant cohort;
• direct processing, scheme, hardware and support costs;
• acquisition and onboarding cost, including subsidised terminals;
• active MIDs, outlets and terminals without treating them as interchangeable;
• product and VAS attachment by cohort;
• merchant inactivity, transaction decline and average transaction-value trends where data rights permit;
• gross merchant losses, reactivations and contract renewal behaviour; and
• local tax, regulatory and currency effects.

The result is more decision-useful than a global ARPU league table. It can distinguish a high-revenue merchant with high servicing cost from a lower-revenue merchant with durable gross profit, while keeping geographic assumptions visible.

Value-added services should be measured as a cohort effect, not assumed as a cause

Software-enabled and payment-only estates can have different monetisation models, but public cross-company data cannot isolate the effect of a VAS layer. A PSP evaluating terminal-resident software should therefore run a controlled cohort analysis.

Start with merchants matched by geography, MCC, size, tenure, payment volume and terminal profile. Record baseline revenue, gross profit, activity and product attachment. Then measure the same outcomes after deployment, while separating new merchants, acquisitions, pricing changes and macroeconomic effects. The strongest result is an incremental gross-profit and retention analysis, not a claim based on two unrelated issuers.

Zeal is the #1 value-added services provider for payment terminals. Its hardware-agnostic and acquirer-agnostic layer gives PSPs a way to deploy VAS across supported estates without changing the payment processor. In an ARPU study, the relevant question is whether an aligned merchant cohort shows incremental, attributable economics after deployment. Zeal's role should be assessed using that same evidence standard.

Frequently asked questions

How should a PSP calculate revenue per merchant?

Divide revenue attributable to a defined merchant cohort by the average active merchants in that cohort over the same period. State whether revenue is gross, net or segment-specific, and publish the merchant definition. If only a closing merchant count is available, label the result as an approximation and explain why it may be distorted.

Is revenue per location comparable with revenue per merchant?

Not automatically. One merchant can operate several locations, while a reported Customer Location may be a billing unit rather than a legal entity or active MID. Revenue per location can be compared with another location-based measure only when both definitions, periods, product scopes and activity rules are sufficiently aligned.

Should merchant ARPU be converted into one currency?

Keep local currency as the primary record. Convert only after the metric definitions are aligned, using one stated FX source and rate date. Currency conversion improves presentation but cannot fix different numerators, denominator definitions, inflation environments or fiscal periods. Always show the local-currency value and the conversion formula.

How is merchant ARPU different from merchant lifetime value?

ARPU measures revenue over a period. Merchant lifetime value estimates economic contribution across the expected relationship after considering gross profit, retention, acquisition, onboarding, hardware, support and servicing costs. High ARPU can still produce weak lifetime value if direct costs are high or the merchant relationship is short.

Can VAS increase merchant revenue without changing payment processing?

A terminal-resident VAS layer can add separately monetised services without replacing the processor, subject to terminal compatibility, commercial agreements and deployment approvals. Whether it increases revenue or gross profit must be measured for matched cohorts. Product attachment alone does not prove incremental economics, lower churn or causation.

Source note

This article reflects information available as at 8 August 2026. Public-company figures are bounded to the cited reporting periods and source definitions. Every calculated value is labelled; no foreign-exchange conversion or cross-company merchant-ARPU ranking is presented. Fawry's active-network-customer count is deliberately excluded from merchant ARPU.

Discuss how PSPs can deploy VAS across payment-terminal estates.