Split funding in a restaurant: the mechanism fits, the margin is the risk
Taking a share of each day's card settlement matches how a restaurant actually collects. That is exactly why it is easy to take too much.
Drafted with AI assistance. Not yet independently checked. Nobody has verified the claims on this page against a source, so treat the figures and legal points as a starting point rather than as settled, and confirm anything you are about to act on. How we check things.
Is split funding a good fit for a restaurant?
Mechanically, yes: a split of daily card settlement tracks how a restaurant collects, falls when you are quiet, and does not bounce a fixed debit on a slow Tuesday. Financially, the risk is that the percentage is quoted against sales while it is paid out of a thin operating margin, so a holdback that sounds small can be most of the profit on the covered volume. Before agreeing, convert the percentage into dollars per week at your real sales level, subtract it from your actual weekly operating profit, and confirm in writing how the holdback can be reduced if sales fall.
Split funding diverts an agreed percentage of each day's card settlement to the funder before the balance reaches you. In a restaurant it is the most honest match between an obligation and a cash flow the market offers, and still the structure most likely to be sized wrong.
Why the mechanism suits the trade
You settle daily, in a fairly narrow band, all year, and a split moves with that. A dead Monday in January remits less than a full Saturday in June, so you are not funding a fixed debit out of a day that did not happen. No returned-item fee for a slow week, no scramble to top up the account before an ACH clears at 6am.
Compare a fixed daily debit, the other common structure. It is easier to model and harder to survive, because it does not know when you are quiet.
Why the margin is the problem
The percentage is quoted against sales. It is paid out of margin. Those are different numbers and the gap between them is the whole risk.
Do that conversion before you talk about price. Percentage of sales into dollars per week, dollars per week against your real operating profit line.
The two clauses that decide how this ends
Stacking is the specific failure mode
Two splits against one settlement stream do not share in any way that helps you. They compound. Add a second holdback to the first and you are remitting the sum of both out of gross card sales, before food, labour or rent — and on the illustrative margin above there is no version of that week that works. An offer of a second advance while the first is live is priced on the assumption that you are already in trouble.
Before you sign
- Convert the holdback to dollars per week at your current volume and at your slowest month's.
- Get the total remittance amount and the expected duration, and ask what the cost looks like if it repays twice as fast. A factor rate has no time dimension, so speed changes the effective cost dramatically without changing the quoted number.
- Confirm the reconciliation mechanism in writing.
- Ask which merchant and depository accounts are covered.
- Ask what happens on a sale of the business, and whether the obligation can be assumed.
Refuse an offer with no reconciliation clause, refuse a stack, and refuse to treat a factor rate as an interest rate. They measure different things and only one accounts for time.
What a true split does when sales fall
The advantage of a genuine split over a fixed debit is that the term stretches instead of the payment failing.
Illustrative only — 100,000 advanced, 130,000 purchased, collected at 10 percent of 60,000 a week of card settlement. That is 6,000 a week and the deal runs about 21.7 weeks.
Card sales fall to 44,000 a week. The split yields 4,400, and the deal now runs about 29.5 weeks — nearly eight weeks longer. The 30,000 of cost has not changed, so the effective annualised cost falls from roughly 72 percent to roughly 53 percent on a simple basis.
That is the honest case for the structure. When trade is bad the product gets cheaper per month and does not bounce. It is also why a funder prefers a fixed debit, and why the word "split" in a sales conversation is worth checking against the document.
Revenue that is not in the split
A split takes a share of what settles through the covered processor. Anything that does not go through it is invisible both to the collection and to the projection behind it.
- Delivery platform payouts, which usually settle from the platform rather than your processor.
- Cash.
- Catering and event deposits paid by cheque or transfer.
- Gift card redemption, depending on how it is processed.
Two consequences. If a meaningful share of revenue sits outside the processor, the deal runs longer than projected — fine for you, and the reason some funders respond by requiring a fixed ACH debit alongside or instead of the split. And if you later move delivery volume onto a channel that does settle through the processor, your remittance rises without anything else changing.
Ask which revenue streams are covered, and get the answer in the document rather than in an email.
The processor is a third party to your deal
The split is performed by your processor under an instruction it has agreed to accept. That makes the processor relationship part of the financing, with consequences people discover late:
- Switching processors, even for a better rate, is usually a default or requires consent.
- A processor that terminates you after a chargeback spike or a risk review breaks the collection mechanism, and the contract's answer to that is generally not in your favour.
- Adding a second processor for a new location or a new channel can be treated as diverting receipts.
If a processing change was on your list, make it before the advance or leave it until the advance is repaid.
Where this applies
Related questions
Is split funding a good fit for a restaurant?
Mechanically, yes: a split of daily card settlement tracks how a restaurant collects, falls when you are quiet, and does not bounce a fixed debit on a slow Tuesday. Financially, the risk is that the percentage is quoted against sales while it is paid out of a thin operating margin, so a holdback that sounds small can be most of the profit on the covered volume. Before agreeing, convert the percentage into dollars per week at your real sales level, subtract it from your actual weekly operating profit, and confirm in writing how the holdback can be reduced if sales fall.
Which funding products does this apply to?
Merchant Cash Advance, Working Capital, Revenue-Based Financing, Credit Card Processing. Each has its own page listing the funders in this directory that offer it and what each one publishes about its terms.
Is this specific to restaurants?
It is written around how a restaurant business actually generates and collects cash, which is what makes its funding problem different. The mechanics transfer; the arithmetic may not.
Who writes this?
The Find Me Funders research desk. Some drafting is AI-assisted, and every page that is says so at the top, including whether a person has checked its claims yet.
How do I know a figure here is right?
Where a page carries the green notice, its claims were checked against the sources listed at the end and a reviewer is named. Where it carries the amber one, nobody has verified it yet and you should confirm anything you plan to act on.
Are the examples real deals?
No. Every worked example is labelled illustrative and exists to show the arithmetic. What any particular lender charges is on that lender's page, where it publishes it.
Why do you never say what a typical rate is?
Because we cannot source it. A market average assembled from lenders who do not publish prices is a guess with a decimal point on it. Where a lender publishes a figure, we show that figure and say where it came from.
Is this financial or legal advice?
No. It is general information about how these products work. Outcomes depend on your contract and your state, and a lawyer or accountant licensed where you are is the person to ask about your situation.
Can I reuse this content?
Quote a paragraph with a link back. Do not republish whole articles.