A cash flow forecast is not a prediction of the future. It is a working document — a set of assumptions you test against reality every month and change when reality disagrees. Get it roughly right and you will spot a lean month while you still have time to do something about it. Get it wrong, or forget to update it, and you will find out about the problem when the bank does.

Profit and cash are not the same thing

You can be profitable and still run out of money. A strong month of invoices raised does nothing for your bank balance until the money actually lands, and for most small businesses and freelancers that gap is where trouble hides.

So build your forecast around cash movements, not sales. Money in is money that has cleared. Money out is money that has left, or that has a fixed date attached to it. That distinction matters most when:

• Clients take 45 or 60 days to pay, whatever the invoice says.
• You have paid out for materials or subcontractors before the client has paid you.
• Quarterly VAT, PAYE, corporation tax or self-assessment payments land in the same month as a quiet trading period.
• You have taken a large deposit that feels like income but is work you still have to deliver.

Start with what you actually know

Open your business bank account and write down today's balance. That is your starting point. Now list the commitments you cannot avoid, with the dates they hit:

• Rent, software subscriptions, insurance, phone and broadband, accountancy fees.
• Payroll, including your own salary or drawings if you take a regular amount.
• PAYE and National Insurance due to HMRC, usually the 22nd of the following month if you pay monthly.
• VAT, due one month and seven days after the end of your VAT quarter.
• Corporation tax, nine months and one day after your year end, plus any self-assessment payments on account on 31 January and 31 July.
• Loan repayments, lease payments, finance agreements, minimum card payments.

These are your baseline outgoings. Add them up for each month. This figure is the number your business has to cover before you pay yourself anything extra.

Estimate income honestly, not hopefully

Most forecasts fail here. Business owners forecast the best plausible month, then are surprised when it does not arrive. Do the opposite.

Go through your pipeline and assign each job two things: the likely value and the likely month the cash will arrive, not the month you will invoice. If a client normally pays late, forecast late. If a contract is not signed, either leave it out or put it in a separate column marked speculative so you can see how much of your forecast depends on it.

Then look at your own history. What did the last twelve months actually look like, month by month? If August is always slow, forecast August slow. If your biggest client takes a two-week shut-down in December, factor it in. Seasonality is not bad luck; it is a repeatable pattern you can plan for.

Add a line for "other" — tax refunds, grants, interest, one-off sales — but keep it small and clearly labelled.

Build a simple month-by-month grid

You do not need specialist software to start. A spreadsheet with twelve columns, one per month, and rows across the left is enough.

• Row 1: opening bank balance.
• Rows 2–4: cash in, split into confirmed, likely and speculative.
• Rows 5–10: cash out, grouped into fixed costs, variable costs, tax, and loan or finance payments.
• Row 11: net movement for the month.
• Row 12: closing bank balance — opening plus net movement.

That closing balance becomes next month's opening balance. Now scan down that final row. Any month that dips towards zero, or below it, is where you need to act. If you want a safety net, aim for the closing balance never to fall below three months of fixed costs. That is the buffer that lets you sleep.

If your income is irregular — and for most freelancers it is — also keep a rolling thirteen-week cash forecast alongside the annual one. Weekly detail catches problems that monthly averages hide.

Compare it against reality every month

A forecast you build once is guesswork. A forecast you update monthly is a management tool. Set aside half an hour on the same day each month, ideally just after your bank reconciliation, and do three things:

• Replace the forecast figures for the month just gone with the actual figures.
• Work out where you were wrong and why — was it a late payer, an unexpected cost, or work that did not come in?
• Adjust the remaining months based on what you have learned.

If you are consistently £2,000 out every month, your assumptions need fixing, not the arithmetic. Track your actual payment days for your top five clients and use those real numbers going forward.

Act on shortfalls before they become emergencies

The whole point of the exercise is to buy yourself time and options. When you see a dip coming two or three months out, you have choices. When it arrives unannounced, you have none.

• Chase invoices earlier. A polite reminder five days before an invoice falls due often brings payment forward by a week or two.
• Delay discretionary spending — new equipment, marketing experiments, that software you have been meaning to trial.
• Negotiate terms with suppliers rather than simply paying late.
• Ask for deposits or stage payments on larger jobs so cash comes in as work progresses.
• Time large purchases into months you know are strong, and keep tax money in a separate account so it never gets spent.

None of this is complicated. It is just consistency: estimate, compare, adjust, repeat. Do that every month and your forecast becomes the most useful page in your business — a quiet early-warning system that tells you what to do next, well before you have to react.