Cash Flow Forecasting for UK Small Businesses — A Practical Template Approach

Why profit and cash are different things

A business can be profitable on paper and still run out of money. If you invoice £10,000 in a month but your customers take 60 days to pay, and your rent, payroll, and supplier bills are due in 30, you can be entirely profitable and still unable to pay a bill on time. Cash flow forecasting exists to catch this gap before it becomes a crisis, not after.

The core structure of a rolling forecast

A useful forecast doesn't need to be complicated. At minimum, it needs three things, projected week by week or month by month for the next 8–13 weeks:

1. Opening balance — what's actually in the bank today
2. Cash in — money you realistically expect to receive, by date, not by invoice date
3. Cash out — every payment obligation you know about, by the date it actually leaves the account

The closing balance for each period becomes next period's opening balance, rolling forward.

Getting "cash in" right

The most common forecasting mistake is treating an invoice date as a cash date. If you invoice on the 1st with 30-day terms, the cash doesn't arrive on the 1st — and in practice, a meaningful share of invoices are paid late regardless of the stated terms. A realistic forecast should:

• Use your actual historical average payment time per customer, not the stated terms, where you have the data
• Separate "confirmed" income (invoiced, contracted) from "expected" income (a quote you're fairly confident will convert) — and weight the expected category down
• Flag your largest customers individually — if one client is 30% of revenue, their payment behaviour alone can swing your whole forecast

Getting "cash out" right

This side is usually more predictable, which is exactly why it's worth getting complete:

• Fixed costs (rent, subscriptions, loan repayments, payroll) — these rarely move and should be entered first
• Variable costs tied to sales (stock purchases, contractor payments, commission) — scale these with your sales forecast
• Tax payments — VAT quarters, Corporation Tax instalments, PAYE, and your own Self Assessment payments on account are all cash-out events that are easy to forget because they don't relate to day-to-day trading
• One-off or irregular costs — equipment purchases, insurance renewals, annual software licences — these cause more surprises than recurring costs because they're easy to forget between renewals

What to do with the output

The forecast is only useful if you act on what it shows. If it flags a shortfall in week 6, the point is to have that warning in week 1 or 2 — giving you time to chase overdue invoices, delay a discretionary purchase, arrange a short-term facility, or have an early conversation with a supplier about payment terms, rather than discovering the gap when a payment actually bounces.

Keeping it realistic

A forecast built once and never updated stops being useful within a few weeks. The habit that actually works is a short weekly or fortnightly update — pull in actual receipts and payments since the last update, compare them against what you'd forecast, and roll the projection forward. Over a few cycles, this also tells you where your own assumptions (customer payment speed, seasonal dips) were wrong, which makes each version more accurate than the last.

This is exactly the kind of forward-looking view that clean, up-to-date bookkeeping makes possible — a forecast built on unreconciled or months-old books is a guess, not a forecast.