Licensing · Essay 02
Royalties — How a Mark Earns Money in Someone Else's Hands
A royalty is only as good as the definition it's measured against.
The contract said five per cent of revenue. Nobody argued about the number — five per cent felt fair to both sides, and the negotiation moved on within an afternoon. Eighteen months later, the licensor discovered that “revenue” in her own contract excluded the licensee's fastest-growing channel entirely, because the clause defined it by reference to the storefront sales the parties had been picturing when they signed, and said nothing about the online sales nobody had pictured yet. Five per cent of a number that had quietly stopped meaning what she thought it meant.
Paul Magaji · 9 min
A royalty is the payment a licence exists to produce, and it is negotiated, almost universally, backwards. Founders fight hard over the percentage — five against seven, seven against ten — and treat the definition underneath it as boilerplate, because a number is concrete and a definition is only words. The number is usually the least important thing in the clause. What is being multiplied matters more than what it is multiplied by, and it is the part almost nobody reads twice.
This essay assumes the base agreement the licence-agreement essay already set out — the reporting obligation and the audit right exist, in outline, before this piece begins. What follows is the part that outline left thin: how a royalty is actually structured, what "revenue" has to say to survive contact with a licensee's accountants, why a minimum guarantee exists, and what happens when the number that arrives each quarter cannot be trusted at face value.
Nobody renegotiates the percentage after signing. Everybody eventually renegotiates the definition.
Act One
Three ways to be paid, and what each one is protecting against
A licence can be priced three ways, and the choice is not cosmetic — each structure exists because of a specific fear, and choosing the wrong one leaves that fear unaddressed.
A lump sum — one payment, or a small number of fixed instalments, unrelated to how the licensee actually performs — protects the licensor against a licensee who will not report honestly, or operates somewhere auditing is impractical. It is simple and it is final, and its cost is that the licensor gets nothing extra if the licensee's use of the name turns out to be extraordinarily successful. She has sold a flat fee for something whose upside she cannot share in.
A running royalty — a percentage of an agreed revenue measure, paid periodically — protects the opposite fear: that a lump sum undervalues a licence that turns out to be worth far more than either party guessed at signing. It aligns the licensor's income with the licensee's actual success, which is exactly why the definition of what is being measured becomes the whole contract.
A hybrid — a running royalty with a minimum guarantee — is what almost every serious licence actually uses, because it takes the running royalty's upside and adds a floor under it: whatever the running royalty produces, the licensor receives at least the guaranteed minimum. Act Three takes the guarantee on its own, because it is doing more work than its one-line description suggests.
Act Two
What “revenue” has to survive
Five mechanisms, each a place where a royalty clause that looked complete on signature turns out to have left something undefined — and each closes on the trap the omission actually sets.
Mechanism One
Gross versus net — and which deductions are real
Every deduction a licensee is permitted to take before the royalty is calculated is a deduction the licensor is agreeing to fund out of her own percentage.
A royalty calculated on gross revenue is measured before any deductions at all. A royalty on net revenue permits stated deductions first — returns, trade discounts, freight, applicable taxes — and the royalty applies only to what is left. Net is the more common structure and is not unreasonable on its face; a licensor should not be paid a percentage of a sale that was later returned and refunded. The clause fails when the list of permitted deductions is left open-ended rather than closed — "and other customary deductions" is an invitation for the licensee's accountants to decide, quarter by quarter, what counts as customary.
The Trap
An open deduction list shrinks silently over time as a licensee's finance team finds new categories to subtract. A closed, enumerated list is the only version of "net" that means the same thing in year five as it did in year one.
Mechanism Two
The channel gap
A revenue definition written around the channel the parties were picturing at signature says nothing about the channel neither party had thought of yet.
The cold open's fastest-growing channel excluded by accident is the common shape of this failure: a definition drafted with retail sales in mind, silent on wholesale, online, marketplace, or export sales that the licensee later adds. The fix is structural rather than exhaustive — define revenue as all amounts received from the exploitation of the licensed mark, however sold, rather than listing channels one by one and hoping the list stays current. A definition anchored to the activity being licensed survives a change of channel. A definition anchored to the channels that existed on the day of signing does not.
The Trap
A licensee under no obligation to volunteer that a new channel exists has no reason to raise the gap in the definition — it is discovered by the licensor, if at all, and usually much later than it should be.
Mechanism Three
Bundling — the royalty base that shrinks by packaging
A licensed product sold inside a bundle is worth its bundled price to the customer and, absent a clause saying otherwise, its allocated price to the royalty.
Where a licensed product is sold as part of a bundle with unlicensed goods — a licensed shoe sold with an unlicensed bag at a combined price — the royalty base should be the licensed product's standalone price, not an allocated fraction of the bundle that the licensee's own pricing team gets to set. Without a clause fixing this, a licensee has a standing incentive to bundle the licensed product with something else and allocate as little of the combined price to it as plausibly defensible.
The Trap
Bundling is rarely dishonest on its face — it is ordinary retail practice — which is exactly why a royalty clause has to fix the allocation method in advance rather than trust that it will be applied fairly after the fact.
Mechanism Four
Related-party sales
A licensee who sells to its own affiliate at a below-market price has not made a sale for royalty purposes — it has moved the point at which a sale is deemed to happen.
Where a licensee sells to an affiliated distributor before the product reaches an independent buyer, the royalty should be calculated on the price the affiliate charges the independent third party, not the internal transfer price between related entities — otherwise a licensee group can route sales through a low-margin affiliate and calculate the royalty on an artificially depressed number, recovering the difference downstream where the licensor's percentage no longer reaches it. A clause requiring related-party transactions to be priced at arm's length for royalty purposes, with the onward sale price substituted where the two diverge, closes this without requiring the licensor to police the licensee's corporate structure directly.
The Trap
This mechanism is the hardest of the five to catch without an audit right specifically extended to related-party transactions — ordinary sales reporting will show the low internal figure and nothing else, by design.
Mechanism Five
Currency and the payment date
A royalty fixed in one currency and paid in another has a second variable riding on top of the sales figure, and the contract should say which party carries it.
Where sales occur in a currency other than the one the royalty is expressed in, the clause should fix the conversion date and source — the rate on the last day of the reporting period, from a named reference rate, rather than whatever rate the licensee finds most convenient to apply. Where the licence itself crosses a border — a Nigerian mark licensed abroad, or the reverse — the deeper currency and repatriation questions belong to a dedicated essay elsewhere in this cluster rather than being absorbed here; what belongs in the base royalty clause is only the conversion mechanics for whichever currency the parties have already chosen.
The Trap
An unfixed conversion date lets a licensee choose, quarter by quarter, whichever historical rate produces the lower royalty — a small distortion each time, compounding over the life of the licence.
Act Three
The minimum guarantee
A minimum guarantee sets a floor: whatever the running royalty produces, the licensor is owed at least the stated minimum for the period. It exists because a running royalty on its own has a failure mode a lump sum doesn't — a licensee who signs a licence, then does not actually try. Exclusivity granted alongside a royalty with no floor costs the licensee nothing to leave dormant; a competitor is kept out of the field, and the licensor's own name earns nothing while it sits unused.
Two design questions decide whether a minimum guarantee does that job or merely looks like it does.
The first is whether the minimum is creditable against the running royalty or payable in addition to it. A creditable minimum means the licensor receives the greater of the minimum or the running royalty for the period — the guarantee is a floor, not a bonus. A non-creditable minimum, paid on top of whatever running royalty is separately earned, is more generous to the licensor and correspondingly rarer to obtain; most negotiated licences use the creditable version, and a licensor should know which one she has rather than assume the more generous reading.
The second is whether the minimum is reviewed against real market performance or fixed for the life of the licence. A minimum set at signing, based on a first-year sales projection, that never adjusts as the licensee's actual market grows, protects the licensor against underperformance in year one and against nothing at all by year five — a running royalty that has grown well past the original minimum makes the floor irrelevant, while a minimum that was never reviewed against a licensee who undershot every year since gives the licensor a number that stopped protecting her long before the term ended. A minimum guarantee reviewed periodically, with a stated mechanism for revision, stays a genuine floor rather than a historical artefact.
A minimum guarantee that has not been renegotiated in five years is not a floor. It is a fact about the year it was signed.
Act Four
Verification — because the number is only ever asserted
Every mechanism in Act Two is invisible without the reporting and audit rights the base agreement already establishes in outline. This is where they earn their keep.
The reporting statement should show its working, not just its total — units sold by category, gross revenue, each deduction claimed and its amount, and the resulting royalty calculation — because a single summary figure gives a licensor nothing to check any of the five mechanisms above against. A statement that shows only "royalty due: [amount]" cannot be audited in any meaningful sense; it can only be believed.
The audit right itself needs three things to be more than a formality: a stated frequency at which it may be exercised without needing special cause, a stated allocation of cost — customarily the licensor pays for a routine audit, and the licensee pays if the audit finds an underpayment above a stated threshold, which gives the licensee a real incentive to report accurately in the first place — and access not just to the licensee's own summary records but to the underlying sales data an independent auditor would need to verify them, including, where related-party sales are in issue, the onward sale prices Act Two's fourth mechanism depends on.
Act Five
Tax, briefly, before the handoff
A royalty paid by a Nigerian licensee, even to a Nigerian licensor with nothing crossing a border, is not simply received in full. Royalty payments are treated as a distinct category under Nigerian tax law and are ordinarily subject to withholding at source, with the licensee deducting tax before the royalty reaches the licensor and remitting it to the tax authority on the licensor's behalf. The applicable rate and treatment shift with policy and, where either party is not Nigerian, with whichever double-taxation treaty may apply — precisely the kind of detail that dates a fixed number the moment it is printed, which is why this essay states the mechanism rather than a rate: a licensor should confirm the current position with an adviser at the time a royalty clause is drafted, not rely on a figure fixed to the year this was written.
Anything the royalty owes to currency conversion, repatriation, or a counterparty outside Nigeria belongs to the essay in this cluster written specifically for licences that cross a border, and is not absorbed here. What belongs in this essay is only the reminder that a royalty figure and the amount that actually lands in the licensor's account are not, even domestically, the same number.
The percentage is the easiest number in the whole arrangement to agree on, and the least important one to get right.
The licensor in the opening did not lose her case for want of a favourable percentage — five per cent was, by any measure, a fair rate. She lost eighteen months of royalty on a channel her own contract had never described, because the number everyone negotiated was multiplying a definition nobody had.
This publication is educational and analytical. It describes how legal and commercial structures work; it does not advise on any particular matter, and nothing here should be relied upon as advice on a reader’s own affairs. The author holds commercial interests in the brand-building and private-label sector examined by this series.