Merchant of record

Who is on the statement, and why that is the whole definition

The merchant of record is the entity whose name appears in the billing descriptor, whose acquiring agreement the transaction is processed under, and which is registered with the card networks as the merchant for that sale. Every other consequence follows mechanically from those three facts, which is why the definition is short and the implications are not.

Because the acquiring agreement is the merchant of record's, the acquirer's rules attach to it: its underwriting, its reserve policy, its settlement schedule, its right to hold funds. Because the network registration is the merchant of record's, the network's monitoring programmes measure it. And because the descriptor is the merchant of record's, the cardholder disputing a charge is disputing against that entity, and the issuer pulls that entity's record.

The contract of sale sits in the same place. The customer buys from the merchant of record, and the terms they accept are the merchant of record's terms. This is the part brands tend to discover late: under a platform merchant-of-record arrangement, the legal relationship at the moment of purchase is between the customer and the platform, and the brand's own terms of sale are downstream of it.

The descriptor has an ordinary commercial cost too. A statement line naming a payments platform rather than the brand is the single largest generator of "I do not recognise this charge" disputes, and it is a brand impression given away at the one moment a customer is looking at what they paid.

Who owes the sales tax and the VAT

This is the substantive reason the merchant-of-record model exists, and it is worth stating precisely rather than as "they handle tax for you".

In most indirect-tax regimes the merchant of record is the seller, and the seller is the party with the obligation. That means registering in each jurisdiction where an obligation arises, charging the correct rate, filing returns on each jurisdiction's calendar, and remitting. In the United States that is economic nexus across state and local jurisdictions with independent rules and thresholds. In the European Union and the United Kingdom it is VAT, with the One Stop Shop and Import One Stop Shop schemes for cross-border business-to-consumer sales. For digital services in many regimes the place of supply is the customer's location, which can create an obligation from the first sale rather than after a threshold.

Many regimes also apply deemed-supplier rules, under which a marketplace or platform facilitating the sale is treated as the supplier for tax purposes regardless of what the commercial contracts say. Where those rules apply, the tax position is not a matter the parties can allocate between themselves.

The value being bought is therefore not compliance work. It is the removal of the registration obligation itself, which matters enormously to a business selling into forty jurisdictions with an entity in none of them, and matters very little to a physical-goods brand selling into one country where it is already registered. The same service is close to essential in the first case and close to worthless in the second, at the same price.

Who owns the disputes, the payouts and the stored credentials

The merchant of record owns the chargeback completely. It is the entity debited, it pays the fee, its dispute ratio takes the count, and its acquirer decides what reserve to impose in response. A brand selling under a platform's merchant-of-record arrangement does not receive chargebacks, because chargebacks are not addressed to it.

From the brand's side that reads as protection, and it is partly real. But the platform passes economic losses back through its agreement, usually by deducting disputed amounts and fees from settlements, and it enforces risk rules on the brand because the brand's disputes land in the platform's portfolio ratio — so it can decline to sell particular products, impose refund policies, delay payouts and terminate the account, with no acquirer relationship for the brand to appeal to. There is a slower cost alongside it: processing history is built by the merchant of record, so a brand that has traded for three years under a platform arrives at its first direct acquiring conversation with no chargeback record, no volume record and no settled disputes, and is underwritten accordingly.

Settlement is the visible difference. A merchant of record settles into its own bank account on the acquirer's schedule; a brand under a platform is paid on the platform's schedule, which is a separate commercial decision and is usually slower. On a growing subscription business the gap is working capital. The fee stack is the one that compounds: a merchant-of-record rate is bundled and prices liability, tax obligation and support alongside processing, so it sits meaningfully above what the same volume would cost on a direct interchange-plus arrangement. The arithmetic is worth running — on two million dollars of annual volume, every one percentage point of bundled-rate premium is twenty thousand dollars a year, and that is the figure to set against what registration, filing and a tax engine actually cost in the jurisdictions you sell into.

The least visible and most consequential is the stored-credential vault. Under a platform merchant-of-record arrangement, the tokens are held by the platform's processor under the platform's acquiring relationship, which means a subscriber file cannot simply be moved. For a subscription business, that single fact can be worth more than every other line in the table below.

Merchant of record versus seller of record

The two terms are used interchangeably and they are not the same thing, which matters as soon as a product causes harm or a regulator asks a question.

The merchant of record is the entity of record for the payment: the descriptor, the acquiring agreement, the network registration, the disputes, and in most regimes the indirect-tax obligation. The seller of record is the entity of record for the goods: named as the seller in the terms of sale, responsible for the accuracy of the description, for fitness and warranty, for consumer-protection obligations, for product liability and for recalls.

In most platform arrangements one party is both, which is why the distinction rarely surfaces. It can be split, and when it is, the consequences are asymmetric: a marketplace may be merchant of record for the payment while the brand remains seller of record for the product, in which case the descriptor and the tax registration move but product liability, safety obligations and consumer-law duties stay exactly where they were.

For cross-border physical goods there is a third role that belongs to neither: the importer of record, the party responsible to customs for duty, import VAT and the accuracy of the declaration. Which party that is follows from the shipping terms — delivered duty paid puts it on the seller, delivered at place typically leaves it with the customer — and it is decided independently of who is merchant of record. A brand can be none of the three, one of them, or all three, and those are different sets of obligations.

When each model is the right trade

There is no default answer, and the honest framing is that the merchant-of-record model buys speed and removes obligations at a price measured in basis points and control. Whether that is a good purchase depends on facts that are specific and knowable.

The two facts that decide it are your tax surface and your volume. Tax surface is how many jurisdictions create an obligation and what registration and filing in each of them would cost you; volume is what the rate premium comes to in cash. Both are computable today, and the model that is right for a brand at launch is frequently the wrong one two years later, which is an argument for choosing with the exit in mind rather than for choosing permanently.

• A platform as merchant of record fits: digital goods or software sold into many VAT and sales-tax jurisdictions; no local entities and no intention to create them; volume low enough that the rate premium is smaller than registration, filing and tax-engine costs; a new market being tested; or a launch that has to happen in weeks rather than months.
• The brand as merchant of record fits: physical goods concentrated in one or a few markets where you are registered anyway; recurring revenue, where credential portability and authorisation-rate control compound over years; volume where basis points on a bundled rate exceed the fully loaded cost of a finance function; a brand that needs its own name on the statement to keep disputes down; and any business that expects to negotiate its own processing terms eventually, because processing history has to start somewhere.
• The move between them is a real project, not a settings change. It means new underwriting, a stored-credential migration that the outgoing provider has to agree to, tax registrations that take time in some jurisdictions, and a descriptor change that will itself generate a wave of unrecognised-charge disputes unless customers are told first.
• Whichever model you choose, ask one question before signing: on the day this relationship ends, what leaves with me? The answer should name the stored credentials, the customer records, the transaction and dispute history, and the terms under which each is exported. A model that cannot answer it is not cheaper, it is just deferred.