How to Build a 13-Week Cash Forecast: Steps, Example, and Weekly Update Routine

To build a 13-week cash forecast, open the opening cash balance, schedule expected customer receipts and scheduled payments in the weeks they will clear the bank, and calculate each week's ending cash as beginning cash plus receipts minus payments. Carry each ending balance into the next week. This guide walks the inputs, the row structure, a worked example, and the weekly update routine FP&A uses to keep the model current.

What is a 13-week cash forecast?

A 13-week cash forecast is a direct-method view of weekly receipts, payments, and ending cash across roughly one quarter. Ending cash equals beginning cash plus expected receipts minus expected payments. The horizon rolls forward one column each week as the oldest week becomes historical actuals.

A monthly close reports last month. An annual budget sets the fiscal-year plan. Neither answers whether Friday's payroll clears or whether to draw on the line of credit for an April tax payment at the weekly resolution a CFO needs mid-quarter. The 13-week forecast fills that gap. Thirteen weeks maps to roughly one quarter: long enough to see a covenant pressure point or a payroll shortfall before it lands, short enough that the weekly assumptions stay inspectable.

What inputs do you need before you build the first column?

Pull three categories of data: bank balances, accounts receivable (A/R) and the collections history, and accounts payable (A/P) and scheduled commitments. Each input needs an owner and a refresh timestamp so the forecast reviewer can see what is current.

Which bank and receivables data should you pull?

Pull these on the same cadence you plan to run the review:

1. Reconciled ending cash balance by operating, trust, and sweep account as of the prior Friday close. The operating balance is the amount available for operating payments. Restricted balances (trust, sweep lockup) belong on their own line. Source from the bank's API or the ERP's bank reconciliation module.
2. Open A/R aging by customer, with invoice date, due date, and expected pay date. Finance owns the expected-pay-date column, which is a view that can differ from the contractual due date for a slow-paying account. Sales ops owns the invoice-date and due-date columns.
3. Days sales outstanding (DSO) trailing 90 days by customer segment. DSO measures the average days from invoice to cash receipt. The forecast's expected receipts rely on DSO being current, not on an annual-budget assumption.

Which payment commitments do you need visibility into?

Four pulls:

1. Open A/P aging by vendor, with invoice date and expected pay date. Treat vendor-set terms as the floor; the controller sets the actual pay date based on cash availability.
2. Payroll run calendar, biweekly or semimonthly, with gross wages, employer taxes, and benefit true-ups by run. Source from the payroll system or the human resources information system (HRIS).
3. Debt service schedule with principal and interest by contractual payment date. Keep covenant test dates on a separate line from the regular payment cadence.
4. One-time items: tax payments, insurance renewals, equipment deposits, severance, deal fees. Each gets a specific week, a specific amount, and a specific owner.

How do you build the forecast step-by-step?

Build the initial forecast in three setup steps, then add scenarios and the weekly update routine.

Step 1: Set the horizon and the weekly columns

Open 14 columns: one historical actuals column plus 13 forecast columns. Label each column with its Friday end-date. The leftmost column holds the completed week's reconciled actuals; the next 13 hold the forecast.

Step 2: Lay out the direct-method row structure

Order the rows so a reviewer can read the week top to bottom:

• Beginning cash balance (operating accounts only)
• Restricted cash balances (shown separately, not available for operating payments)
• Customer receipts, broken out by top-10 customer and "all other"
• Other receipts (tax refunds, financing draws, dispositions)
• Payroll and benefits
• Supplier payments
• Debt service (principal and interest, by contractual payment date)
• Tax payments
• Capex and discretionary
• Other payments
• Ending cash balance
• Operating minimum (the internal floor finance will not take operating cash below)
• Cash above operating minimum (ending cash minus the operating minimum)
• Covenant requirements (shown separately: contractual minimums tested on contractual dates)

The "cash above operating minimum" row is the one the CFO reads first. When it approaches or crosses zero in any forecast week, that is the trigger for a funding conversation. Covenant minimums are tracked separately so an internal buffer decision does not get confused with a contractual requirement.

Step 3: Populate the forecast weeks

Enter the reconciled opening cash balance in Week 1's beginning-cash row. For each of the 13 forecast weeks:

• Place expected customer receipts in the week they are expected to clear the bank, using the expected-pay-date column from the A/R file.
• Place scheduled payments (A/P, payroll, debt, tax, capex) in their scheduled weeks.
• Calculate ending cash as beginning cash plus total receipts minus total payments.
• Carry that ending balance into the next week's beginning cash row.

Below is an illustrative two-week slice of the opening calculation, in thousands:

The Week 1 figures are illustrative, not a Centage benchmark. A negative "cash above minimum" in Week 1 is the signal the forecast is built to surface.

Step 4: Add scenario toggles for the exposures that could change a cash decision

Choose scenarios around exposures that would change a cash decision inside the horizon: a slower-paying anchor customer, a delayed financing close, a tariff or input-cost spike, a payroll change. For each scenario, record which receipt or payment row changes, by how much, and in which week. The scenario count should match the business's actual risk profile and the review's time budget rather than a fixed number.

Step 5: Save the approved forecast as a baseline

Save the forecast as the approved baseline before the first weekly update. The baseline is what you compare next week's actuals against; without a saved baseline, variance analysis has no reference point.

How do you update the forecast each week?

Update the forecast weekly so changes in expected receipts and payments reach the next cash review before a decision needs to be made. One combined routine covers the actuals load, the variance review, and the forward update.

Each week, in order:

1. Load actuals. Replace the oldest forecast column with the completed week's reconciled receipts, payments, and ending cash. Move the reporting window forward one column so the horizon stays 13 weeks forward.
2. Calculate variance by row. Signed variance is actual minus forecast. For a nonzero forecast row, percentage variance is (actual minus forecast) divided by the absolute forecast amount. When the forecast is zero and actuals are not, report the dollar variance only. Agree with the forecast reviewer on which variance size prompts a one-line note.
3. Record the cause of any material variance. Note the amount, the cause, and whether the receipt or payment moved to a different week in the horizon. Prioritize variances that affect cash above the operating minimum or a pending cash decision.
4. Revise expected dates and amounts in the forward 13 weeks. When an assumption changes, update the affected weeks and record the update in a change log.
5. Review cash above the operating minimum across the horizon. If any week approaches or crosses zero, name the two actions (draw, defer, collect) that close the gap.
6. Record the reviewer, the sign-off time, and the single cash decision this update informs. The sign-off field lives in the forecast file.

Updating weekly keeps the forecast predictive. Shifting to a monthly cadence turns it into a reporting view and removes the mid-quarter decision support that is the point of the 13-week horizon.

Why do 13-week forecasts in a spreadsheet drift over time?

Three failure modes to check during the weekly update:

Each failure mode has a mechanical fix inside the model, and each fix compounds the weekly maintenance load. The deeper issue is that a spreadsheet tracks generic rows rather than the drivers that move cash in your business. For how that driver architecture varies by industry, see cash flow forecasting for mid-market CFOs: driver architecture across five industries.

How does Centage build the 13-week forecast with your team?

Centage does not sell a template. Our team of AI FP&A experts builds the forecast around how the business actually runs: chart of accounts, entity structure, payroll calendar, top customers' payment terms, debt service, and covenants. The forecast lives inside Centage as a connective layer across the general ledger, payroll, pipeline, operational systems, and bank data, so DSO, payroll, and debt service refresh from the systems that own them rather than from a weekly manual export.

That connective layer is what reduces the weekly maintenance load the three failure modes above create inside a shared spreadsheet. If you want the FP&A function run alongside your team by a US-based advisor, Managed FP&A adds that layer; both the software and the Managed FP&A engagement are priced during discovery against your entity count, chart-of-accounts depth, and reforecast cadence.

Want to discuss your 13-week cash forecasting process with Centage? Book a demo. Tell us which systems hold your actuals today, how you update the forecast each week, and the cash decision the next review needs to answer.

Frequently asked questions

Can we build a 13-week forecast without pulling directly from the ERP?

Start with reconciled bank balances, open receivables, open payables, and scheduled commitments exported from the source systems. Give each input an owner and a refresh timestamp. Direct connections to the ERP, HRIS, and bank reduce weekly manual work and remove a class of transcription errors; whether to add them is a payback calculation against the hours the manual exports consume.

Should the 13-week forecast replace the monthly reforecast?

No. The 13-week model answers whether cash clears across the quarter at a weekly resolution. The monthly reforecast updates P&L and balance-sheet projections for the fiscal year. Build both, and keep the 13-week coupled to the monthly reforecast's assumption set so the two do not diverge.

How many scenarios should the forecast carry?

Start with scenarios covering the exposures that could change a cash decision inside the horizon, such as delayed collections, financing timing, and a supplier-payment acceleration. Pick the count that matches the business's risk profile and the time the review can give each scenario. Refresh the scenario list quarterly, or sooner when a new exposure emerges.

How many weeks forward can we trust the forecast?

Near-term weeks (roughly the first month) depend on A/R aging, A/P aging, and the payroll calendar, which are the inputs the forecast reviewer can refresh directly. Mid-horizon weeks depend more on DSO and expected pay dates, which carry more judgment. Weeks beyond that are directional and move most on scenario toggles. If the near-term variances are consistently large, the inputs are stale before the model is.