Cash flow forecasts are the most useful spreadsheet a small business owner ever builds – and the most ignored. The reason most forecasts get abandoned isn’t complexity; it’s that they’re built once, never updated, and quickly become wrong. This guide walks through how to build a 13-week rolling forecast you’ll actually maintain, how to test it against scenarios, and the assumptions that catch out new founders.
What a cash flow forecast actually does
A cash flow forecast predicts the cash in your bank account week by week. It surfaces upcoming crunches before they happen. It lets you stress-test decisions – “if I hire now, when does cash run low?”. It gives banks and lenders confidence when you need finance. None of it works if it isn’t maintained – a forecast last updated three months ago is worse than no forecast, because it creates false confidence.
The 13-week rolling forecast — why this window
Long enough to see meaningful issues. Short enough to be accurate. Aligns with quarterly business rhythms (VAT, payroll, key supplier terms). Updated weekly so accuracy doesn’t decay. The 13-week window has been the small-business standard for decades because it balances effort and value better than any other length. Anything shorter is reactive; anything longer is fiction.
Inputs you need
Opening bank balance. Customer invoices outstanding and when payment is expected. Forecast new sales – signed but not yet invoiced. Recurring revenue – retainers, subscriptions. Confirmed expenses – rent, salaries, suppliers. Tax payments – VAT, Corporation Tax, PAYE. Any one-off planned expenses – equipment, training, capital purchases. For most small businesses an hour collecting these inputs is the bulk of the work.
Step-by-step build
- Open a spreadsheet with 13 weekly columns plus a labels column.
- Row 1: opening cash balance.
- Rows 2–10: cash in by source (invoices due, recurring revenue, new sales weighted by stage).
- Rows 11–25: cash out by category (salaries, rent, suppliers, tax, capex, marketing).
- Row 26: net cash movement (cash in minus cash out).
- Row 27: closing balance (opening + net movement).
- Carry closing balance forward as next week’s opening.
- Highlight any week where closing is below your safety threshold (1 month of expenses, typically).
A worked example: 4 weeks of a real forecast
Concrete numbers help. A small consultancy starts a Monday with a £14,000 opening balance. Week 1: two retainer invoices land (£3,500 each = £7,000); cash out covers £4,200 of salary, £600 rent, £400 software, £180 marketing. Net movement +£1,620; closing balance £15,620. Week 2: a project milestone payment arrives (£8,000); cash out is salary-light (£800 contractor day-rate), £200 expenses. Net movement +£7,000; closing balance £22,620. Week 3: quarterly VAT payment of £6,800 hits, plus salary £4,200, plus £1,500 supplier invoice. Cash in: £3,500 retainer. Net movement -£9,000; closing balance £13,620. Week 4: a slow week on cash in (only £1,200 from a small invoice); cash out £1,000 software renewal and £800 marketing. Net movement -£600; closing balance £13,020.
Three lessons from these four weeks. The VAT payment in week 3 is the single biggest cash event of the month and is forecastable months ahead – the cash crunch it would cause is entirely preventable with a separate VAT savings account. Retainer revenue is the smoothest line on the forecast and the easiest to predict accurately. Project milestone revenue is lumpy: week 2 looks great, but if the milestone slips to week 4, the closing balance dips below the founder’s comfort threshold. The forecast surfaces all three before they happen.
Common categories of cash in
Customer invoice payments per existing debtor. Recurring revenue – retainer, subscription. New sales forecast based on pipeline. Tax refunds – VAT reclaim, Corporation Tax overpayment. Director’s loan in (if you’re funding the company). Sale of assets. Investment or loan drawdown. The discipline of separating each source matters: it lets you spot which line of cash is unreliable.
Common categories of cash out
Salaries, dividends, drawings. Rent and utilities. Tools and subscriptions. Marketing. Suppliers and contractors. Tax (VAT, Corporation Tax, PAYE). Loan repayments. Capital purchases. Be honest about each – the most common forecasting error is forgetting recurring expenses that don’t move the needle individually but add up to 10–20% of cash out.
Seasonality, deposits and milestone payments
Three patterns trip up otherwise-careful forecasts. Seasonality: if your business has a recurring quiet quarter (summer for B2B services, January for hospitality), the forecast must show the dip rather than smooth across it. Look at last year’s actuals by week and bake the pattern in. Deposits: when you take 30–50% upfront on new client work, the deposit is cash in now against costs you’ll incur over the next 4–8 weeks – treat the deposit as a liability internally, not as freely-spendable income, otherwise you’ll spend it before the work is delivered. Milestone payments: a £15,000 contract paid in three £5,000 milestones is three separate cash events, each at risk of slipping a week or two. Forecast each milestone explicitly, with the client’s typical approval lag built in. A consultancy I know learnt the hard way that “invoice on milestone” in practice meant invoice 5 working days after milestone, then 30 days for payment. That’s 7 weeks from milestone date to cash in bank, not the “net 30” on the contract.
Scenario planning
Base case: what you expect to happen. Optimistic: 20% upside on sales, faster collections. Pessimistic: a major client loses, payments slow, an unforeseen cost lands. For each, run the model and identify the cash low point. The pessimistic scenario tells you how big a buffer you actually need – usually it surprises founders by being bigger than their gut said.
Assumptions that bite you
Customers will pay on time (they often don’t). Sales pipeline closes at expected rate (optimistic). One-off costs are one-off (rarely). Tax bills are fixed (they’re not – late penalties, interest, surcharges all bite). New clients pay 50% upfront (sometimes negotiated away). Retainers won’t churn (they will). Stress-test against each of these regularly; the buffer you need always exceeds the buffer you think you need.
Maintaining the forecast
Weekly: 30 minutes Monday morning, update with actuals. Monthly: review the next 13 weeks fresh – re-forecast based on what you’ve learned. Quarterly: compare actuals to forecast retrospectively. What were you systematically wrong about? Adjust your assumptions. The discipline of looking back at where the forecast missed is what makes the next forecast more accurate.
Frequently asked questions
The questions UK small businesses ask most often about cash flow forecasting.
How accurate should a cash flow forecast be?
For the first 4 weeks, aim for 90%+ accuracy on cash out and 70-80% on cash in. For weeks 5-13, accuracy drops to 50-70%. Beyond 13 weeks, forecasting is more strategy than prediction. The point isn't precision — it's seeing the shape of the next quarter.
What's the difference between cash flow and cash flow forecast?
Cash flow is what actually happened (last quarter's bank movements). Cash flow forecast is what you predict will happen (next quarter). Both matter: actuals teach you about your business; forecasts help you steer it.
Should I include things I might do but haven't decided?
Generally no, in the base case. Build the base case on commitments you've actually made. Use the optimistic and pessimistic scenarios to play with 'what ifs'. This keeps the base case usable as a decision tool rather than a wishlist.
How do I forecast new sales when I don't know what'll close?
Weight your pipeline by stage. A signed contract = 100%. A verbal yes = 70%. A second meeting = 40%. A first contact = 10%. Multiply each opportunity's value by its probability and aggregate. As you close more deals, refine the percentages based on your actual close rate.
What about VAT in a cash flow forecast?
Track gross (VAT-inclusive) cash in and cash out, with a separate line for VAT due to HMRC each quarter. Don't try to net VAT out at the line-item level — it gets confusing fast. The quarterly VAT payment goes on the cash-out side as a single big number.
How far ahead should I forecast?
13 weeks rolling for tactical decisions. 12 months high-level for annual planning. 3 years for big strategic decisions or fund-raising. Don't try to do a weekly 12-month forecast — the precision is fake.
My forecast says I'll run out of cash in week 9. What do I do?
Three options, in order: (a) accelerate cash in (chase debtors, push for upfront from new clients), (b) defer cash out (negotiate supplier terms, delay non-essential spend), (c) inject cash (director's loan, business overdraft, invoice factoring). Most cash crunches resolve with options (a) and (b) given 6-8 weeks of notice — the forecast's purpose is to give you that notice.
Should I share my cash flow forecast with anyone?
With your accountant and (for limited companies) your fellow directors, yes. With investors or lenders if you're raising. With suppliers offering payment terms, sometimes — a credible forecast can extend the terms they're willing to offer. Not with employees outside the senior team, generally.
