There is a specific and very common kind of panic: the year has gone well, the work is booked, and there is not enough money in the account on Thursday. Nothing has gone wrong. Profit and cash are simply not the same number, and the gap between them is where small businesses fail.
Profit is what you earned. Cash is what has arrived. An invoice raised in October for work delivered in September, paid in December, counts as a good September and an empty November. Every business has this gap. The ones that survive know how wide theirs is.
Why a profitable business runs out of money
Three mechanics produce almost all of it, and none of them is a mistake you made.
- Your costs are monthly and your income is lumpy. Software, rent and anybody on your payroll leave on the same days regardless of whether a client paid.
- Growth eats cash. A bigger project means more hours worked before the first invoice goes out, so the better the month you are having, the further ahead of your income you are working.
- Tax arrives in instalments that do not care about your calendar. The money was never yours, but it sat in your account long enough to feel like it was.
The thirteen-week forecast
You do not need accounting software to fix this. You need one sheet with thirteen columns — a quarter, week by week — and four rows. Thirteen weeks because it is far enough out to give you time to act and near enough that you can still guess honestly.
| Row | Where the figure comes from |
|---|---|
| Opening balance | Your actual bank balance on Monday. Not the accounting figure. |
| Money in | Unpaid invoices, placed in the week you realistically expect payment |
| Money out | Standing costs, plus anything you have already committed to |
| Closing balance | Opening plus in, minus out — and this is the only cell that matters |
Then read the closing-balance row across and find the lowest number. That single figure is your cash-flow position: how close you come to the floor and in which week. Everything else on the sheet exists to produce it.
The four levers, fastest first
When the lowest number in that row is uncomfortable, there are exactly four things you can do about it. They are not equally fast.
- Invoice sooner. The fastest and the most overlooked, because it costs nothing and changes today. Anything delivered but not invoiced is money you have decided to lend for free.
- Take deposits. Moves cash to the start of a project instead of the end, and doubles as a client filter — see the post on late payers.
- Bill in stages. Three invoices across a project instead of one at the end shortens the gap between working and being paid without asking anyone to pay faster.
- Shorten terms, or slow payments out. The slowest lever: existing terms are already agreed and renegotiating them takes a cycle to land.
Notice what is not on the list: chasing harder. That helps with a specific late invoice, and it does nothing about the structure that made you dependent on it. Chasing is a symptom-level fix and the first three items above are structural.
How much buffer is enough
The honest answer is a number of months, not an amount, because an amount stops meaning anything the moment your costs change. Three months of fixed costs is the figure most people converge on — enough to absorb losing your largest client without making a decision in a week.
Getting there is boring: a fixed percentage of every payment received moves to a separate account on the day it arrives, and the buffer is not a place you borrow from for a good opportunity. If it funds opportunities, it is not a buffer, it is your current account with an extra step.
Three numbers, once a month
You do not need a dashboard. You need three figures on the first of the month, and you need to be able to see whether each is getting better or worse.
- Days to payment, per client. Invoice date to money-in-bank, averaged. Terms tell you nothing; this tells you everything, and it names the clients quietly financing themselves with your cash.
- Total unpaid, split by how overdue. One number is comfort, four buckets are information. The over-60 bucket is the one that turns into a write-off.
- Weeks of runway. Bank balance divided by weekly fixed costs. The one figure worth knowing without looking it up.
Only the first of those requires anything you might not have. Days to payment needs the invoice date and the payment date in the same place, which is the whole argument for having invoicing and payment matching in one system rather than an invoice tool and a bank tab. Reconstructing it by hand from statements is possible once and never twice.
Days-to-payment is only free if the dates live together
When invoices are issued and payments matched in the same place, the payment history per client is a by-product rather than a project. That is what makes the forecast above take minutes instead of an afternoon.