It was a Friday afternoon in our Tangier office and I was looking at a payroll run I could cover, but only just. Fifteen salaries. Everyone had worked a full month. The money meant to pay them was supposed to have arrived four weeks earlier from a single client, a retainer I had celebrated signing that spring because it was the biggest one we had landed all year.
The work had been delivered. The reports had been approved in writing. The invoice had been acknowledged with a friendly note saying it was in the queue.
Eleven weeks was the eventual gap between the month we did the work and the day the transfer cleared. Nobody defaulted. Nobody disputed a line item. No lawyer was ever involved. The client simply paid the way they paid everybody, slowly, and I had signed a twelve-month agreement without ever asking a single question about how they paid.
The Risk That Never Shows Up in the Forecast
Every forecast I had built until that point made one silent assumption: that an invoice becomes cash on the date printed on it. That assumption is doing enormous work, and nobody stress-tests it, because it feels like accounting rather than risk.
It is not accounting. It is the single largest exposure most small service firms carry, and it is entirely invisible on a revenue chart. The Federal Reserve Banks' 2025 Report on Employer Firms found that 56% of small employer firms named paying operating expenses as a financial challenge and 51% named uneven cash flows, sitting just behind rising costs. Those two numbers describe the same wound from opposite sides. Money leaves on a fixed schedule and arrives on a flexible one.
Unpaid client invoices and a calculator on a finance desk, representing accounts receivable risk in a service business. Photo by Jakub Żerdzicki on Unsplash.
For a firm that sells inventory, a slow payer is annoying. You can stop shipping. For a firm that sells people's time, the cost of the month is already gone by the time the invoice is issued. Salaries were paid. Software was billed. The strategist and the two specialists spent February on that account and February cannot be returned to the shelf.
What Three Years of Invoices Actually Showed Me
That weekend I did something I should have done years earlier. I exported every invoice we had issued since founding the company and added one column: days between issue date and the date the money actually cleared our account.
The average across the whole book was 38 days. That felt fine. Then I broke it out by client and the average stopped meaning anything.
Two clients out of the eleven active at the time were sitting at 71 and 84 days. Everyone else was between 12 and 30. Those two accounts were not slightly worse than the rest. They were a different business model quietly attached to ours, one where we financed their working capital for free while carrying all the delivery risk ourselves.
The one that ran to 84 days was on a $6,500 a month retainer. Over a year, that is $78,000 of revenue, and at any given moment we were carrying roughly $18,000 of it as unpaid work already delivered. We had, without ever deciding to, become that company's cheapest lender.
The Screen I Run Before Signing Now
It takes about twenty minutes and it happens before the proposal, not after the contract. I ask four things:
- Who physically releases the payment, and how many people touch it first? A founder with a bank app pays in four days. A finance department with a three-signature approval chain and a monthly payment run pays in sixty, no matter what the contract says. This one question predicts more than everything else combined.
- Will they pay a deposit? Not because I need the cash, but because a company that resists a 30% deposit on a first engagement is telling you exactly how the rest of the year will feel. Two prospects have walked away at this question. Both times I was relieved within six months.
- What does month one actually cost us to deliver? Onboarding-heavy engagements front-load our cost into the exact period where we have the least evidence about how this client behaves. If the ramp is expensive, the terms need to be shorter, not longer.
- How much of one month's payroll would this account represent if it stopped paying tomorrow? That converts an abstract worry into a number I can look at without flinching.
None of this requires a finance background. The broader thinking behind it lines up with what Harvard Business Review's financial management coverage and the US Small Business Administration's guidance on managing business finances both keep circling back to, which is that liquidity and profit are separate questions and a business can pass one while failing the other.
Founder reviewing a spreadsheet of invoice payment dates to measure how long each client actually takes to pay. Photo by Ruthson Zimmerman on Unsplash.
Why Service Firms Get Hit Harder Than They Expect
When a digital marketing agency commits a strategist, a paid media specialist, and a designer to a monthly retainer, the margin on that account is decided by the twelfth of the month, long before anyone thinks about collections. That is why late payment lands differently in a digital marketing agency like ours than it does in a firm that can slow production down. Our capacity is a perishable good. An unbilled week of a senior specialist's time does not roll forward to next month, it evaporates, and the client is still owed the same deliverables.
We work across Morocco, Dubai, and the United States, and the pattern holds in all three markets with different accents. In the UAE, payment cycles on corporate accounts routinely run 60 to 90 days and everyone treats that as normal, so the risk is not that a client is bad, it is that you priced and staffed as though they were fast. In the US, terms are usually cleaner but the distance makes chasing awkward. In Morocco, invoices tied to a public or semi-public entity can sit in an approval chain for a full quarter with nothing wrong at any step.
What Actually Changed
Four things, and none of them were dramatic.
We now hold a reserve equal to two months of fixed costs, and I stopped counting it as available money. We charge a deposit on every new engagement, which was uncomfortable to introduce and turned out to bother almost nobody. Our internal forecast now shows an expected-cash line beside the revenue line, built on each client's own historical payment lag rather than on invoice dates, and those two lines being visibly different is the whole point. And every retainer over a certain size carries a written pause clause: if an invoice passes 45 days, work stops until it clears, stated plainly at signature rather than deployed as a threat later.
That last one felt aggressive to write down. In practice, it has been invoked twice in two years, and on both occasions the client apologized and paid within a week, because the pause was never a surprise. The clause did not make anyone pay faster out of fear. It made the expectation legible, which is a different and much more durable thing.
The client from that Friday afternoon is still with us, incidentally. They still pay slowly. The difference is that I now price for it, staff for it, and know exactly what it costs, which means it stopped being a risk and became a line item. Risk is only dangerous while it is unnamed.
If your revenue chart looks healthy but the bank balance keeps telling a different story, the gap is usually sitting in payment behavior nobody screened for. That is one of the things we dig into with founders at Rhillane Marketing Digital, the digital marketing agency I run across Morocco, Dubai, and the United States.