The 13-week cash flow forecast, done properly.

Why thirteen weeks

A twelve-month forecast is a plan; a thirteen-week forecast is a warning system. Beyond a quarter, most UK SMEs cannot say with any confidence when a customer will actually pay, and a forecast built on guesses trains you to ignore it. Inside four weeks you already know — the question is weeks five to thirteen, which is exactly the window in which a cash problem is still cheap to fix.

The rows that matter

Keep the shape simple. One line per category, one column per week, and never more categories than you will actually maintain.

• Opening cash — the real cleared balance on Monday morning, not the ledger balance.
• Receipts — grouped by customer, dated by when the money arrives, not when the invoice goes out.
• Payroll and pension — fixed amounts on fixed dates, including employer NI and any auto-enrolment uplift.
• Rent, insurance, software and finance leases — the recurring items you tend to forget in the third month.
• VAT and Corporation Tax — take the dates from your own HMRC account rather than assuming quarterlies; interest and penalties for late payment accrue from the day the bill is due.
• Supplier payments — grouped by payment terms, with the ones you genuinely chase shown separately.
• Owner drawings and loan repayments — the line that quietly empties an account.
• Closing cash — opening plus in minus out. That is the only number anyone reads.

How to fill in receipts without lying to yourself

Take every open sales invoice and put it in the week the money historically lands for that customer, not the week it is due. Most ledgers will tell you that a customer who is due on 30 days pays on 47. Use that number. If a customer has never paid before, treat the receipt as absent until it clears, and if it is large, decide now what you do if it does not arrive.

The weekly rhythm

The forecast is worth nothing if it is a document you rebuild monthly. Fifteen minutes, same time every week: roll the window forward, move last week's actuals in, change any assumption that reality has contradicted, and read the lowest closing balance in the next thirteen weeks. That single number is what you are managing.

• Monday: enter last week's actual receipts and payments.
• Tuesday: chase the three largest overdue invoices — the forecast assumes you tried.
• End of month: compare forecast to actual and adjust the slippage you keep discovering.

What breaks forecasts

Three things, every time: paying a supplier early because they asked, treating a deposit as revenue, and letting the tax line float because the bill has not arrived yet. A deposit is a liability until you deliver the work. The tax bill is certain even when the amount is not, so put in a placeholder on the due date and correct it when the return goes in.