TLDR: We hire a permanent role only after the revenue that pays for it has been billed and collected for three consecutive months, collected being the word that matters. The year we hired on a signed contract instead, the client paid late and then shrank the scope, and two salaries ran for five months on cash the business had mostly not received. A vetted freelancer bench now covers the gap between a signature and the third collected invoice, and it has kept us from repeating the mistake.
The signature came through on a Tuesday afternoon and I hired two people before the end of the week. I remember the feeling exactly: relief that the pipeline had finally converted, and a kind of impatience to staff the work before the client changed its mind. The contract was the largest we had signed at the time, a twelve-month engagement across search and paid media for a group with operations in Morocco and the Gulf, and it justified, on paper, two permanent roles.
The first invoice went out on schedule. It was paid seven weeks later. The second was paid nine weeks after that, and by then the client had quietly reduced the scope by a third, citing an internal reorganisation. The two people I had hired were excellent. They were also being paid every month from cash that had not arrived, and I was the one who had put them there.
The Problem With Hiring on Signed Revenue
Service businesses hire against contracts because contracts are the only forward-looking number they have. A signature looks like certainty. In a small agency it is closer to an intention: the client intends to pay, on the terms written, for the scope described. Any of those three can move.
Payment terms move first. Across our client base, which bills in Moroccan dirhams, UAE dirhams and euros or US dollars depending on the client, the distance between an invoice date and a bank credit varies by market and by client far more than the contract suggests. Scope moves second, usually downward, usually with a good reason. And the intention itself can change with one departure on the client side.
The JPMorgan Chase Institute's report on cash flows, balances and buffer days measured how few days of cash the typical small business actually holds against its outflows. When I first read it, I did the sum for my own company and realised that two new salaries had eaten most of our buffer before the first invoice was paid. The contract was real. The cash was not there yet, and a salary does not wait for it.
What It Cost
The arithmetic was not complicated, which made it worse. Two salaries for five months, against roughly one and a half months of collected revenue from the contract they were hired for. The difference came out of runway, and runway in a small service business is measured in the months you can keep paying everyone if nothing new is signed.
We lost about four months of it, not to a bad client and not to bad work, but to the timing gap between a signature and collected cash, a gap I had chosen to ignore because the signature felt so good.
Nobody was let go. We covered the shortfall by pulling forward other invoices, by my own salary going to zero for a quarter, and by an unglamorous stretch of cutting every tool and subscription that was not producing revenue that month. I also stopped taking on new engagements that would have needed more staff, which is the hidden cost: a cash-constrained agency turns down growth precisely when it needs it.
The Rule
The rule that came out of that year has one sentence and one hard word in it.
A permanent role is opened only after the revenue that pays for it has been billed and collected for three consecutive months.
Collected, not billed. Billed revenue is a promise; collected revenue is a bank line. Three months, because one collected invoice can be a client paying its deposit before going quiet, and two can be a client paying properly before reducing scope. Three consecutive collections mean the engagement has a rhythm and the client has a habit. Consecutive, because a gap resets the count. If the client pays months one and three but skips two, we are looking at a cash pattern, not a reliable revenue line.
The rule applies to the role, not to the person. If a freelancer we already trust wants the permanent seat, the same three months apply. If a new contract would fund a role we already need for other clients, the three months still apply, but the count can start from the existing clients' collections rather than the new one's.
Harvard Business Review's argument that your approach to hiring is all wrong is mostly about how companies choose people. The part I keep coming back to is its point that most firms hire without measuring whether the hire paid off. Our rule is a blunt way of measuring before the hire, with the one number that cannot be argued with.
What Covers the Gap
A rule like that has an obvious weakness. Clients do not wait three months for work to begin, and a signature that cannot be staffed is worth nothing. The answer is a freelancer bench, and it has become one of the more deliberately managed parts of the business.
The bench is a short list of specialists we have already worked with, paid a small monthly retainer to be reachable, with agreed day rates and a 48-hour activation window. When a contract is signed, the first three months of delivery are staffed from the bench, at a cost that scales with the work and can be switched off if the scope shrinks or an invoice is late. If the third month collects, the role is opened and the bench freelancer gets the first offer. If it does not, nobody has been hired into a salary the business cannot pay.
HBR's research on thriving in the gig economy is written from the freelancer's side, and it is worth reading from that angle: the people on our bench value the predictability of a retainer and a known activation process far more than I expected, and it is what keeps them answering when we call. The bench has an internal cost too. Freelance day rates are higher than the equivalent salary, and briefing a freelancer takes more of an account lead's time than briefing a colleague. We accept both, because the alternative is what happened the year we skipped it.
Where the Rule Lives Now
The rule is written into how we run a digital marketing agency like ours: every open role in our planning sheet carries the name of the revenue line that funds it and the dates of the three collections that justified it. If those cells are empty, the role stays a bench assignment, whatever the signed pipeline says.
I still feel the same impatience when a large contract comes in. The difference is that the impatience now goes into activating the bench by Thursday, which is a good use of it, instead of into two offer letters, which was not.
If You Run a Service Business
Write down the gap, in days, between your last ten invoice dates and the dates the cash actually landed, by client and by currency if you bill in more than one. Then look at your last three hires and ask which of them were made inside that gap. If the answer is most of them, the rule will feel severe. It felt severe to me the first year. It has felt like a floor ever since.