13-Week Cash Flow Forecast: A Practical Guide for UK SMEs

Profit is an opinion; cash is a fact. A UK SME can post a healthy trading result and still miss payroll because a large customer moved to 75-day terms and a VAT quarter landed in the same week. The 13-week cash flow forecast exists to make that collision visible before it happens.

What a 13-week forecast actually is

It is a direct-method model: opening bank balance, forecast receipts, forecast payments, closing bank balance — repeated across thirteen consecutive weeks. Every figure traces back to a real invoice, a contractual commitment or a statutory deadline. Nothing in it is derived from the P&L.

That distinction matters. Accruals-based reporting deliberately smooths timing. Liquidity management is entirely about timing, so the forecast has to ignore the accruals and follow the bank account.

The line items to include

• Receipts from customers — driven from the AR ledger, aged by expected payment week rather than invoice due date.
• Other receipts — R&D tax credit, grant tranches, VAT refunds, asset disposals, investor drawdowns.
• Payroll and PAYE/NIC — net pay on the pay date, PAYE and NIC on the 22nd of the following month.
• Supplier payments — from the AP ledger, grouped by payment run rather than spread evenly.
• VAT — payable one calendar month and seven days after the quarter end.
• Corporation tax — nine months and one day after year end for companies below the large-company threshold.
• Debt service and leases — capital, interest and any covenant-linked sweeps.

UK statutory timing that breaks most models

The two most common errors in a UK 13-week forecast are both timing errors, and both are avoidable.

Building it, step by step

1. Fix the opening balance to the actual cleared bank position, not the ledger balance.
2. Export the AR ledger and assign each open invoice an expected payment week based on that customer's historic behaviour, not their stated terms.
3. Export the AP ledger and map each open bill to the payment run you actually intend to include it in.
4. Overlay payroll, PAYE, VAT, pension and debt service on their statutory or contractual dates.
5. Add recurring operating costs — rent, software, insurance — on their real direct-debit dates.
6. Model the funding lines separately: undrawn facility, invoice discounting availability, overdraft limit.
7. Sanity-check week one against the bank feed. If week one is wrong, nothing after it is credible.

The weekly variance review is the product

A forecast that is never compared to actuals is a spreadsheet, not a control. The discipline is a fifteen-minute Monday review: last week forecast versus actual, largest three variances by value, cause of each, and whether the cause is one-off or systemic. Systemic variances get pushed into the assumptions; one-offs get noted and forgotten.

After six or seven cycles the model stops being a guess. That is the point at which it becomes useful in a board meeting or a lender conversation, because you can evidence the accuracy rather than assert it.

Scenarios worth keeping alongside the base case

• Your largest customer pays 30 days later than forecast.
• A planned funding tranche slips one quarter.
• Payroll grows by the hires already in the plan, with a two-month lag to revenue.
• A downside where receipts run at 85% of forecast for six consecutive weeks.

Each of these should answer one question: in which week does the closing balance breach the minimum operating cash level? That week, not the average, is what you manage against.

Doing this in MouCFO

MouCFO builds the 13-week grid directly from your AR and AP ledgers, applies UK VAT, PAYE and corporation tax timing automatically from your financial year-end and VAT quarter, and lets you override any individual week. Forecast versus actual variance is tracked each week, and the resulting liquidity narrative flows straight into the board pack.