A 13-week cash-flow forecast is not a miniature annual budget. It is a weekly operating control: what cash is available, what is expected to arrive, what must leave, and which decisions happen before the low point.
Short answer: begin with reconciled bank cash, forecast receipts by customer or collection behavior, schedule disbursements by due date, separate committed from discretionary cash, and review forecast-versus-actual every week. The value is not the thirteenth week. It is the decision made in week three.
Why thirteen weeks works
Thirteen weeks is long enough to see payroll cycles, rent, debt service, tax dates, major vendor payments, and the cash effect of current sales. It is short enough that a manager can name the transactions instead of hiding uncertainty inside annual percentages.
Nothing makes thirteen mathematically sacred. A highly seasonal or project business may need more visibility. The discipline is weekly receipts and disbursements, not the number on the tab.
Step 1: reconcile opening cash
Start with cash that can actually be used.
Separate:
- Operating bank accounts.
- Payroll or tax accounts.
- Restricted cash.
- Undeposited receipts.
- Outstanding checks or pending electronic payments.
- Credit availability, shown separately from cash.
Do not type the general-ledger balance into the forecast and move on. Tie it to the bank and explain the difference. Every later week depends on the opening number.
Step 2: forecast receipts from behavior
Revenue is not a receipt schedule.
For card sales, use the settlement pattern and fee timing. For invoices, use open receivables and actual collection behavior. For deposits, use signed work and expected billing milestones. For recurring customers, separate contracted, scheduled, and merely expected revenue.
A useful confidence structure is:
| Receipt class | Meaning |
|---|---|
| Committed | Invoiced, scheduled settlement, or contractually due |
| Probable | Supported by repeat behavior or work already underway |
| Possible | Pipeline or management expectation without a collection event |
Keep “possible” receipts visible but out of the base case until the operating evidence supports them.
Step 3: schedule disbursements by obligation
Build from records, not memory:
- Open accounts payable.
- Payroll calendar and payroll taxes.
- Rent, utilities, insurance, and software.
- Debt principal and interest.
- Materials, subcontractors, and freight.
- Sales and income tax dates.
- Capital spending.
- Owner draws or distributions.
Separate the date an invoice was recorded from the date cash will leave. Then mark each payment as committed, timing-flexible, or discretionary. “Flexible” does not mean optional; it means management can negotiate or choose the week.
Step 4: calculate the weekly roll-forward
The core schedule is simple:
beginning cash + receipts − disbursements = ending cash
The next week begins with the prior week’s ending cash. Add a control row that checks this roll-forward for every column.
Illustrative example
| Week 1 | Week 2 | Week 3 | |
|---|---|---|---|
| Beginning cash | $82,000 | $71,000 | $29,000 |
| Receipts | 64,000 | 48,000 | 76,000 |
| Disbursements | (75,000) | (90,000) | (61,000) |
| Ending cash | $71,000 | $29,000 | $44,000 |
The three-week total looks survivable. Week two is the decision point. The owner has time to accelerate a collection, stage a vendor payment, delay discretionary capital spending, or draw an already-approved credit line. A monthly model could hide that low point.
These figures are illustrative and not a client result.
Step 5: show minimum cash and available liquidity
Management needs a boundary. Set an operating minimum that reflects payroll, critical vendors, and the practical volatility of receipts. Then show:
- Ending cash.
- Less operating minimum.
- Excess or shortfall.
- Undrawn committed credit, separately.
- Liquidity after the credit line.
Do not count an unapproved loan, hoped-for equity, or a credit line already near its covenant limit as available cash.
Step 6: attach decisions to the low points
A cash forecast without decisions becomes a weather report.
For each week below the required threshold, name:
- The driver of the shortfall.
- The action available.
- The owner of the action.
- The date by which it must happen.
- The fallback if it does not.
“Collect faster” is not an action. “Owner calls the three invoices over $10,000 by Tuesday and confirms payment dates” is.
Step 7: run three narrow cases
Avoid duplicating the entire workbook into Base, Upside, and Downside tabs. Test the few variables that matter:
- Collection dates slip by one or two weeks.
- Sales or bookings fall below plan.
- Gross margin declines.
- A hiring or capital decision moves earlier.
- One large customer or project is delayed.
The output should answer: what is the lowest cash week, how much is the gap, and how early must management act?
Step 8: review variance every week
Freeze last week’s forecast before updating it. Then compare:
- Forecast receipts versus actual receipts.
- Forecast disbursements versus actual disbursements.
- Timing differences versus permanent changes.
- New information that management should have known earlier.
Repeatedly optimistic receipts are not a spreadsheet problem. They are a forecasting behavior that should change the next assumption.
Common failure modes
Using the P&L as cash. Revenue and expense recognition do not determine bank timing.
Ignoring taxes and debt principal. Both consume cash even when the P&L presentation differs.
Hiding uncertainty. One number for a speculative receipt makes the schedule look more certain than the operation.
Updating history. Overwriting last week’s forecast destroys the variance record and prevents improvement.
Balancing with borrowing. If debt automatically appears whenever cash goes negative, the model assumes away lender approval, covenant capacity, and timing.
The owner’s weekly page
The detailed schedule can contain hundreds of lines. The owner page needs five:
- Current unrestricted cash.
- Lowest projected cash and the week it occurs.
- Receipts at risk.
- Payments requiring a decision.
- Actions due before the next review.
That is enough to run the meeting. The detail exists so each number can be challenged.
A 13-week forecast is most useful when it sits beside a line-level P&L read, because margin and working capital can move in opposite directions. If the underlying operating drivers are unclear, the financial-model review standard shows how source reconciliation, scenarios, and cash controls are tested.