Nobody argues about the percentage
A royalty is arithmetic: a rate applied to a base. The rate is in the agreement, both parties signed it, and in years of operating a network you will rarely see it questioned.
The base is a different matter entirely. It is a number that one party computes and the other party pays on — reported monthly, by people running their own businesses, from systems the franchisor did not choose, on a definition that may or may not have been read since signing.
So the entire weight of the commercial relationship rests on a figure that is produced once a month by somebody else and taken largely on trust. That is a strange thing to leave unexamined, and most networks do.
This page is about that figure: what it needs to contain, how it goes wrong innocently, how it goes wrong otherwise, and what can be checked without turning a partnership into an inspection.
What we do not do
We do not calculate your royalties
The rate, the treatment and the invoicing are yours. We make the base complete, comparable and traceable — which is the part that involves reading documents.
We do not interpret your agreement
What counts as the base, what audit rights exist and how disputes resolve are contractual questions for your legal advisers. Nothing here is advice on any of that.
We do not accuse anybody
Software can surface a pattern worth asking about. It cannot conclude that a number is wrong, and a tool that claimed to would be doing damage in a business built on long relationships.
We do not fill a gap
A missing submission stays visible as missing. Estimating it produces a number that hardens into fact, which is a specific and avoidable disaster described further down.
What counts as the base
“Gross sales” sounds unambiguous and is not. Here are the questions two honest parties can answer differently from the same month's trading.
| Question | Both answers are defensible |
|---|---|
| Before or after refunds? | Refunds are not sales — or the customer did buy and later returned |
| Including VAT? | It is money taken — or it was never the operator's money |
| Delivery platform orders? | The customer bought your product — or the operator only ever received the net |
| Gift vouchers: sale or redemption? | Cash came in then — or the trade happened later |
| Staff meals and comped items? | No money moved — or the product left the building |
| Sales tax on top in some regions? | Consistency across the network — or consistency with local practice |
Every one of these has been the subject of a real dispute somewhere. The agreement usually addresses two or three of them and is silent on the rest — often because the format that raises the question, delivery platforms being the obvious example, did not exist when the template was drafted.
The practical consequence is that a franchisor should be able to say, for each operator, which convention that operator is using. Not assume it is the same as everyone else's, and not discover it during a dispute.
Making that explicit and carrying it alongside the figures is the same problem as location rollup solves for owned estates, and it matters more here because there is money moving between two parties on the back of it.
Read a period of submissions now
Upload what your operators actually send — summaries, till reports, spreadsheets, photographs of printouts — and see them come back as rows. Free tier, no registration.
Upload a document
Drop your documents here
or browse to upload
Free: 10 pages / month · Upgrade for more
Files encrypted in transit · auto-deleted after processing · GDPR compliant
The fields a royalty run needs
| Field | Why |
|---|---|
| Operator and site | One operator may run several sites on different terms |
| Period covered | The month the trading happened in |
| Date submitted | So lateness and low sales stop looking identical |
| Net, VAT and gross separately | Whichever the agreement specifies is then available without reopening documents |
| Refunds | The most common definitional difference, and invisible if netted off |
| Platform sales separately | The second most common, and growing |
| Basis note | Which convention this operator reports on, recorded once |
| Source file and page | So a query is answered in a minute rather than by asking them to resend |
The last row does disproportionate work in a franchise setting. Asking an operator to resend a report from fourteen months ago is a small request that lands badly — it reads as being checked up on. Answering the same question from your own record is invisible to them.
Forty operators, forty formats
This is the defining practical feature of franchise reporting and the one that distinguishes it from an owned estate. Franchisees are independent businesses. They buy their own equipment, choose their own accountants and run their own back office.
So a monthly submission might be a PDF from a till system, a spreadsheet with a personal layout, an export from an accounting package, a photograph of a printed Z-report, or a figure typed into an email. All of them are legitimate; none of them match.
The instinct is to mandate a template. Networks try this regularly, and it works about as well as any mandate on people who own their own businesses: the diligent comply, the busy send what they always sent, and within a year there are two formats instead of one.
Reading whatever arrives is the more robust approach precisely because it needs no compliance. It also has a quieter benefit — an operator who is not being asked to change their admin is an operator with one less reason for friction, and in a network friction compounds.
The operator who reports late
Lateness is the most common problem in royalty collection and the most commonly mishandled, because it gets merged with the figure itself.
If a period has no submission, the royalty run for that period shows nothing. Merged into a total, that looks exactly like an operator who sold nothing — and the responses to those two situations could not be more different.
Keeping the period date and the submission date as separate fields fixes it completely. A gap is then visibly a gap in reporting, not a gap in trading, and it can be chased as an admin matter rather than escalated as a commercial one.
It also gives you something more useful than a chase list: a lateness pattern. An operator who is consistently ten days late has an administrative problem, and it is usually solvable with a conversation about how they produce the report. An operator who suddenly goes from punctual to three weeks late has changed something, and that is worth a different conversation.
Estimating a missing period, and why it hardens
When a submission does not arrive, the temptation is to estimate — take last month, or the same month last year, and invoice on that so the cycle is not held up.
It is a reasonable-sounding decision that causes a specific and repeatable problem. The estimate goes onto an invoice. The invoice gets paid. Neither side treats the matter as open any more, and when the real figure eventually arrives — if anyone still asks for it — reconciling it means reopening a settled month. Most of the time nobody does, and the estimate has quietly become the record.
The direction of the error is not random either. Estimates get anchored on the last known figure, and an operator whose trade is growing has every incentive not to correct one that was too low.
Two habits keep this clean. Mark an estimated period as estimated in the data, not just in an email. And settle it explicitly when the real figure arrives, even if the difference is small — because the habit of leaving estimates unsettled is what turns one convenient shortcut into a systematic understatement.
This is why a missing submission is left visible here and never filled. A gap is uncomfortable and gets chased; a plausible number is comfortable and does not.
Under-reporting, and what it actually looks like
Worth addressing directly, and worth being careful about, because the assumption does more damage than the problem in most networks.
The overwhelming majority of base discrepancies are definitional or administrative. An operator excluding platform sales because they only ever see the net, or netting refunds because that is what their report shows, is not concealing anything — they are reporting what their system puts in front of them.
Deliberate under-reporting exists, and what it looks like is worth knowing so it can be distinguished from the innocent version.
A series that is too smooth
Real trading is lumpy — weather, holidays, local events. A site reporting within two percent of itself every month for a year is reporting a habit, not a business.
A step change with no cause
Reported sales drop eight percent in one month and stay there, while transaction counts and orders from you do not move.
Divergence from your own data
Product ordered from the franchisor keeps rising while reported sales flatten. That is the most useful signal a franchisor has, because it comes from your side.
A tender mix that shifts
Card proportion falling steadily in a market where it is rising everywhere else.
None of these prove anything. Each is a reason to ask a question, and in most cases the answer is mundane — a new supplier arrangement, a change in how the report is produced, a local employer closing.
The third is the strongest because it uses information the franchisor already holds and the operator cannot shape. It is also the one most often not looked at, because supply data and royalty data usually live in different systems and nobody has put them side by side.
Four checks worth running every period
Completeness before anything else
Every operator, every period, present or explicitly marked absent. A royalty total computed on an incomplete base is wrong in a direction that flatters nobody.
Each operator against its own last year
The same argument as any multi-site comparison: it removes size, location and format differences automatically.
Reported sales against product supplied
Your own data on one side of the comparison. The single most informative check available to a franchisor.
Basis consistency
Has this operator's convention changed? A quiet shift from gross to net is a real event and looks like a modest decline.
The first is not glamorous and it is where most of the money is. Networks routinely invoice a period with two or three submissions missing, chase them for a while, and then move on — and those periods are never billed at all.
When an audit clause gets used
Most franchise agreements contain a right to examine an operator's records. It is rarely exercised, and the reasons are commercial rather than legal: invoking it changes a working relationship into an adversarial one, sometimes permanently.
What that means in practice is that the clause is not really a routine control. It is a last resort, and by the time it is used the relationship has usually already deteriorated.
The consequence is worth stating plainly: the ordinary monthly data is the actual control. If the base is complete, comparable and traceable every month, most problems surface as questions long before anyone reaches for the contract.
And if the clause does get used, the quality of your own records determines how it goes. A franchisor who can produce every submission, its arrival date and its stated basis for four years is in a very different position from one who has an inbox.
It is a relationship, not a tax return
A point that is easy to lose when the subject is data quality. A franchisee is not a subsidiary and not a supplier — they are a business owner who chose your system and can, in the long run, choose to leave it.
That shapes what good practice looks like here. Every additional reporting requirement is a cost you impose on someone whose goodwill you need. Every accusation that turns out to be a definitional misunderstanding is withdrawn far more slowly than it was made.
So the design principle is: take on the burden yourself. Read whatever they send rather than mandating a format. Answer your own historic queries rather than asking them to resend. Ask about a pattern before characterising it.
There is a self-interested version of the same argument, too. A network where reporting is easy gets reported to; a network where it is a chore gets late, thin submissions — and late thin submissions are exactly the conditions under which a real problem stays hidden.
The moment to fix all of this is at signing
Everything on this page becomes easier or harder depending on decisions taken in the first month of a franchise relationship, when nobody is thinking about royalty data because they are thinking about opening.
Agree the basis explicitly, in writing, per operator
Not 'gross sales' but the four specifics: refunds, platforms, VAT, staff and comped items. Ten minutes at signing, versus a dispute in year three.
Establish what their equipment produces
Ask for a sample report before the first submission is due. It is a neutral request at that point and an intrusive one later.
Set the submission channel
One destination, not a person's inbox. Personnel change and inboxes do not transfer.
Explain what happens if a period is missed
Agreed in advance, this is administration. Improvised during the first miss, it is a confrontation.
The asymmetry is stark. Each of these costs minutes at the start of a relationship and is nearly impossible to introduce later without the operator wondering what prompted it.
For an established network the equivalent moment is a renewal or a transfer. Both are natural points to confirm the basis without it reading as a change of posture.
Legacy operators and inconsistent terms
Any network more than a few years old has operators on historic terms — a different rate, a different definition, an exemption someone granted in 2018, a promotional period that was supposed to end.
This is normal and not a problem in itself. It becomes a problem when the terms live in the agreements and the royalty run assumes uniformity, because the run then quietly applies the wrong basis to two or three operators every single month.
The practical remedy is to carry the terms as data rather than as documents. Which rate, which basis, from which date, for this operator — held where the calculation can see it rather than in a filing cabinet.
It also has a side benefit worth having: it makes the drift visible. Networks are often surprised, when they lay it out, by how many variants exist and how few of them were deliberate policy rather than an accumulation of individual accommodations.
Six traps
Assuming every operator uses the same definition of gross sales. Two or three of them will not, and it will be discovered during a dispute.
Merging the period date with the submission date. Late reporting and low trading then look identical.
Estimating a missing period and never settling it. The estimate becomes the record, and the error has a direction.
Invoicing on an incomplete base. Missing submissions get chased for a fortnight and then never billed at all.
Comparing operators to each other rather than to themselves. It measures the estate, not the operator.
Treating a pattern as a finding. In most cases the explanation is mundane, and the accusation is not retractable.
Where to start
Take one completed period and every operator's submission for it — including the ones that arrived as photographs and the one that is always a figure in an email. Read them all into one table.
Then do the unglamorous part: for each operator, write down which basis they report on. Refunds in or out, platforms in or out, VAT in or out. One line per operator, kept permanently.
For most networks that afternoon turns up two or three operators whose basis is not what head office assumed — and it is far better to find that now, in a routine exercise, than during a disagreement about a specific invoice.
After that the checks are cheap and repeatable. Completeness, own-history, supply comparison. Whole-network comparison is on how to compare performance across locations, and the operator's own side of this is on finance for franchise operators.