Thirteen weeks, and why profitable businesses run out of money

Profit and cash are different things

A business can be profitable and still fail to make payroll, and this catches people who are otherwise good at their trade.

Profit is a statement about a period: revenue earned minus costs incurred. Cash is a statement about a moment: what is in the account today. They diverge because of timing. You buy materials in March, do the work in April, invoice at the end of April on 30-day terms, and get paid in June. Three months of outflow before the inflow, on a job that is profitable.

Growth makes this worse, which is the part that surprises people. More work means more materials bought before more invoices are paid. The businesses that run out of cash are frequently the ones doing well.

The instrument

A thirteen-week rolling cash forecast. One page, one column per week, three blocks.

• Opening balance. What is actually in the account.
• In. Invoices due, by the week you expect them rather than the week they are due. Add expected new sales.
• Out. Wages, rent, suppliers, tax, loan repayments, subscriptions, the quarterly bill you always forget.
• Closing balance, which becomes next week’s opening.

Thirteen weeks because it is far enough to see a problem while you can still act (chase early, delay an order, arrange a facility, talk to the tax authority before rather than after) and near enough to be estimable.

The row that matters is the closing balance. Any week where it goes negative is a date you now know about in advance, and knowing it in advance is the entire value.

Where AI helps, and where it must not

Helps: building the structure. Give me the layout and the formulas for a thirteen-week rolling cash forecast in Google Sheets, with opening balance, receipts, payments and closing balance. Explain each formula. Twenty minutes instead of an afternoon.

Helps: listing what you forgot. Here are my payment categories. What does a small [trade] business typically pay that is missing from this list? Insurance renewals, annual licences, tax instalments, the accountant’s fee. The forgotten items are what break a forecast.

Helps: reading it. Paste the numbers and ask which week is tight and what the largest lever is. It will usually name the obvious thing, and the obvious thing is usually right.

Must not: the numbers. Every figure is yours, from your bank, your invoices, your contracts. A forecast with an estimated figure in it is a forecast about a business that does not exist.

Must not: the arithmetic. Formulas in the sheet, tested on three rows.

The habit

Fifteen minutes, same time every week. Update the opening balance from the bank, move anything that did not happen, add what is new.

Doing it weekly matters more than doing it precisely. A rough forecast updated every Monday is a working instrument; a careful one built in January and never touched is a document.

The two questions it answers

Can I afford this? A new machine, a hire, a van. Put it in the relevant week and look at the closing balance line. This converts an anxious feeling into a visible answer.

Which invoice do I chase first? Not the oldest, and not the largest. The one whose arrival changes a negative week. That is a different list from the one you would have chased on instinct, and it is the whole reason to have the forecast in front of you when you write the chasers.

If you do only one thing

List every payment leaving your account in the next four weeks, with dates, and compare the total against your balance plus confirmed receipts.

Half of small businesses have never written that down. It takes thirty minutes, needs no software beyond a sheet of paper, and it is the single most useful thing in this module.

Estimating the uncertain rows

Two rows are always guesses: new sales, and when customers will actually pay. Guess them in a way that helps.

Pay from history, not from terms. If your customers pay on average 47 days against 30-day terms, forecast 47. Your accounting software can tell you the real figure in a minute, and using the terms instead is the single most common reason a forecast is optimistic.

Forecast new sales low. Not out of pessimism, but because the cost of being wrong is asymmetric. Overestimating cash you do not have leads to commitments you cannot meet; underestimating leaves you pleasantly surprised.

Show the bad case as its own line. One extra row at the bottom: closing balance if the two largest receipts arrive a month late. If that line stays positive, you can stop worrying about it. If it does not, you know precisely which two invoices decide your quarter.

Tell the truth early

If a week is going to be short, the useful actions are all things that need notice: talking to the supplier about terms, asking the tax authority about an instalment arrangement, chasing the two invoices that matter, delaying an order.

All of them work far better three weeks ahead than three days. The forecast is worth having only because it buys that notice.