Why 13 weeks
Long enough to see big payments coming. Short enough to be accurate. Standard among lenders and investors.
The build
Step 1 — Opening cash position
Pull the actual combined cash balance across every business account on Monday of week 1.
Step 2 — Receipts schedule
- Confirmed: signed POs, direct debits, recurring subscriptions
- Probable: invoices issued, paying on average DSO
- Pipeline: weighted by close probability and DSO
Step 3 — Payments schedule
- Payroll, PAYE, pension (fixed dates)
- Rent, utilities, insurance (fixed dates)
- Supplier bills (use historical DPO)
- Tax: VAT, corporation tax (use deadline dates)
- Capex: planned only
Step 4 — Net position by week
Opening cash + receipts − payments = closing cash → next week's opening.
Discipline
- Update weekly, on the same day, by the same person
- Lock the prior week as "actuals"
- Roll forward — add a new week 13 each time
- Track forecast accuracy: % variance week 1 → 4
Scenario layers
- Base — what you actually expect
- Downside — top customer pays 30 days late
- Stress — top customer doesn't pay at all
If any scenario goes negative, plan the response now: invoice finance, overdraft, accelerated receipts, deferred capex.
Tooling
- Float — best-in-class, syncs live with Xero/QuickBooks
- Fluidly / Futrli — AI-assisted projections
- A clean Excel model is fine if maintained weekly
Ernest & Co Premium clients get a fortnightly cashflow review with a senior accountant.