Cash flow, liquidity, runway: why a profitable month can still leave you short

A freelancer, on the twentieth of the month. Over the last three months the business invoiced 18,000 and spent 12,000. Of that, 6,000 has been collected and 12,000 is still owed by clients. There is 800 in the account and 3,000 of bills falls due before the fifth.

Every one of those four numbers is correct. Three of them are comfortable and the fourth decides what happens on Friday. This is not an unusual business or a badly run one. It is an ordinary freelancer with good clients who pay slowly, and it is the single most common way a working business gets into trouble.

The confusion is not carelessness either. Profit is the number everybody is taught to look at, it is the one an accountant reports, and it is the one that sounds like success. It is also the only one of the four that can stay high while the business fails.

The reason the words are separate is the reason the example above works. A business is not one thing that is going well or badly. It is four things, and they can point in different directions on the same morning.

Of the four, runway is the only one that tells you what to do. Profit tells you the work was worth doing. Cash flow tells you what happened. Liquidity tells you what you hold. Runway tells you how much time you have, and time is the thing decisions are made out of.

It is one division. Take the cash you actually hold, take what leaves in a normal month, and divide. In the example above: 3,000 a month leaving is 100 a day, and 800 in the account is eight days.

Say it in days rather than months, because months round away the part that matters. “Under a month” sounds survivable. Eight days is a decision. The same arithmetic, expressed honestly, behaves completely differently in a human head, and that is not a trick, it is the point of measuring anything.

Three things make a freelancer’s runway different from the version a funded company talks about, and all three make it easier to get wrong.

So the honest version is narrower than the flattering one: cash you hold, minus what was never yours, divided by what leaves in a day. That number is usually smaller than people expect the first time they compute it, and it is the first number in this whole subject that changes behaviour rather than mood.

What it changes is the question you are asking. Without it you ask whether things are going well, which has no answer and no action attached. With it you ask whether you can afford to turn down the badly-priced job, wait out the slow client, take the week off, or buy the equipment. Those have answers.

Everything above computes from information you already have. No software is required to start and no accounting knowledge is assumed. Three lines, one page.

There is a shorter version, used by credit committees to decide whether a company can pay them this month: short-term receivables, minus short-term liabilities, plus the cash in the bank. Inventory is deliberately left out, because goods in a warehouse have never repaid anybody on time. It is a blunter instrument than the three lines, and it takes ten seconds.

This is the line that decides whether the sheet is useful or decorative. Your invoice says thirty days. That client pays in forty-seven. Forty-seven is the real number, and your rent depends on the real number.

Almost every article about getting paid is about setting terms. Terms are a wish. What a client has actually done, repeatedly, is a fact, and it is sitting in data you already hold: the date you issued each invoice, and the date the money arrived.

Two other things can make the amount that arrives differ from the amount you invoiced, and both look like a short payment until you check: exchange rate differences and bank charges on the way. And when an invoice genuinely is late, what you are owed and what to do is a page of its own.

One correction before the sheet can be trusted. The cash line is not all yours. If you charge VAT, part of that balance belongs to the state and is simply passing through. The tax on your profit has not been assessed yet and is coming. Contributions arrive on a schedule you do not control.

A runway computed on the raw balance is therefore optimistic by exactly the amount you will least enjoy discovering. Take the set-aside out first, then compute: how much to set aside, and which of it was never yours.

Start the sheet by hand. It is the right way to begin, it costs nothing, and it teaches you what the numbers mean in a way no software will. What it will not do is stay true.

A runway is only worth anything if it is current. Yesterday’s runway is not a smaller version of today’s, it is a different number about a business that no longer exists in that state. And a number you rebuild by hand is current on the morning you rebuild it and drifting by the afternoon: an invoice goes out, a payment lands short because a bank took its cut, a subscription renews, a client who used to pay in thirty days quietly moves to fifty.

That is the whole difference between a tool and a system, and it has nothing to do with features. A tool performs a task you asked for. An invoicing tool tells you what you sent. An accounting package tells you, accurately and later, what already happened. Neither answers the question you actually woke up with, which is whether you can survive this month, because that question needs several numbers to be true at the same moment.

A control system is simply the arrangement where the answer is standing there when you look, assembled from things you were doing anyway. You issue invoices, you record costs, money arrives. Those events already contain the runway. What a system does is refuse to let them sit in four places that never meet.

None of this requires prediction, and you should be suspicious of anything that claims it. The useful version is unglamorous: arithmetic on facts you already have, kept current, shown as one number, with the tax reserve taken out before it is shown to you.

The value of this is not accuracy, which it will never fully have. It is rhythm. A number you look at every week tells you about a change while it is still small and still cheap. The same number computed in a panic tells you about a change that already happened.

So: same day, same ten minutes, before the inbox sets the agenda. What lands, what leaves, how long you last. If it takes longer than ten minutes, the sheet is too detailed, and a sheet too detailed to read weekly is a sheet you will stop reading.