Most small businesses that fail were profitable on paper at the time. They ran out of cash because the money went out on a schedule and came in on a slower one, and nobody was watching the gap on a weekly basis. A rolling 13-week forecast exists to make that gap visible roughly a quarter ahead of the day it matters.
Why profit and cash are not the same thing
Your profit and loss statement follows accounting rules about when income and costs are recognised. Your bank account follows a much simpler rule: money either arrived or it did not. The difference between the two lives in three places:
- Timing. You invoice in March, the customer pays in May, your supplier wants paying in April.
- Stock and work in progress. Money converted into inventory is still your money, but it is not available to pay wages.
- Lumpy items. Tax, insurance and annual subscriptions arrive as single large hits rather than smooth monthly costs.
The only three columns you need
You do not need accounting software to build this. A spreadsheet with one row per week and three columns is enough to start:
- Money in, by expected date. Customer payments, deposits, refunds you are owed, grant or loan instalments if they are certain. Use the date you expect the money to clear, then quietly add a few days of buffer to any payment that depends on someone else's process.
- Money out, by required date. Wages, rent, supplier invoices, loan repayments, card fees, subscriptions, tax. Include the small ones; they add up and they cluster.
- Opening and closing balance. Start with your real bank balance today, add the week's money in, subtract the week's money out, and carry the closing balance into the next week.
That is the whole model. Columns to the right of those are commentary, and commentary is optional until the forecast is honest.
Building it in an hour
| Minutes | Step | Notes that save you time later |
|---|---|---|
| 0–5 | Open your banking app and write down today's real balance | One account at a time. If you have several, add a row for each and total them. |
| 5–20 | List every payment you are owed, with the date you expect it and the customer name | Split large invoices across two dates if the customer has a history of paying late. |
| 20–40 | List every payment you must make, on its due date | Include tax, card fees, insurance and anything annual. These are the lines people forget. |
| 40–50 | Lay them out across 13 weekly columns and calculate the running balance | Weeks run from your update day, not from the 1st of the month. |
| 50–60 | Mark the lowest closing balance and write the date next to it | This is your decision date, not the date the forecast ends. |
Read the low point, not the closing balance
The number that matters is the lowest closing balance across the 13 weeks, and how far away it is. Three questions turn that number into a decision:
- Is the low point positive? If yes, you have room. If it is negative, the gap is real and you have a date to work towards.
- How many weeks away is it? Ten weeks gives you options: negotiate terms, slow a purchase, chase invoices. Two weeks gives you almost none.
- What has to go right for that number to hold? If the low point depends on one customer paying on time, treat the low point as the optimistic case and model a late payment.
Five levers when the low point is negative
| Lever | How fast it moves cash | The trade-off to accept |
|---|---|---|
| Chase and invoice faster | Days to two weeks | Requires a consistent process and pleasant persistence; costs you time, not money |
| Shorten payment terms or charge a deposit | Two weeks to a month | Some customers will push back; a deposit for new customers is normal in most industries |
| Negotiate longer terms or staged payments with suppliers | Immediate if agreed, slow if refused | Ask before the invoice is overdue, not after — goodwill is the asset you are spending |
| Delay or reduce a planned purchase | Immediate | Watch for contracts already signed; cancelling those is a cost, not a saving |
| Raise prices on new work | Weeks, compounding | Slowest lever and the most durable — see the pricing playbook |
Notice what is not on the list: borrowing to cover a shortfall you have not diagnosed. Debt is a legitimate tool, but it converts a timing problem into a fixed monthly obligation. Use it after you know which of the five levers above you are actually pulling, and how much of the gap remains once you have pulled them.
Make it a ten-minute weekly habit
- Same day every week. Monday morning or Friday afternoon — whichever day you can genuinely protect.
- Replace the week that has passed with actual numbers. Did that invoice get paid? Did the supplies cost what you thought?
- Update the cash balance from your bank, not from your forecast. The forecast is a plan; the bank is the truth.
- Add anything new — a quote you sent, a purchase you agreed, a tax date you were reminded of.
- Write the new low point and its date at the top of the sheet. If it moved further out, you are winning. If it moved closer, you have this week's agenda.
Frequently asked questions
What is the difference between cash flow and profit?
Profit is what is left after costs in a period, regardless of when money moves. Cash flow is what actually lands in and leaves your bank account. A sale on 60-day terms is profit today and cash in two months, which is why a growing, profitable business can still miss payroll.
Why 13 weeks rather than 12 months?
Thirteen weeks is roughly one quarter — long enough to see a funding gap forming and short enough that the weekly detail stays accurate. A 12-month forecast is useful for planning and useless for deciding what to do next Tuesday.
How often should the forecast be updated?
Once a week, on the same day, and it should take about ten minutes. Update the opening balance with the real figure from your bank, replace the week that just passed with actuals, and check whether the low point moved.
Should tax payments be included?
Yes, as separate lines dated on the day they are actually due. Late tax is one of the few business debts that escalates quickly and is difficult to renegotiate, so it belongs in the forecast with its own date rather than lumped into an overhead estimate.