The 13-Week Cash Flow Forecast, Explained

The 13-week cash forecast is the single most useful management report a growing company can run, and the least glamorous. It is a grid: thirteen weekly columns, rows for cash in and cash out, a running balance along the bottom. That is the whole tool. What it buys you is the difference between finding a cash problem eleven weeks out — when every fix is cheap — and finding it the Thursday before payroll, when none are.

Why thirteen weeks

One quarter. Short enough that you can forecast from actual known items — invoices already sent, bills already received, payroll already scheduled — rather than from assumptions. Long enough to see around the corner: a slow collections month in week three shows up as a balance problem in week nine, while you still have six weeks to act.

Monthly forecasts hide the danger inside averages. A month can look fine in total while week two of it dips below zero — payroll on the 15th, big receivable landing the 22nd. Cash problems happen on days, not months.

What goes in the rows

Receipts, by name. Each significant expected payment as its own line, with the week you actually expect it — not the week the invoice says. Your real collection lag is knowable from history; use it. Recurring card receipts can be one modeled line.

Disbursements, by commitment. Payroll (largest and least movable), rent, debt service, insurance, tax payments — including the quarterly estimates and filings that ambush companies every April and June; pull dates from the Tennessee filing calendar so they are never surprises. Then the discretionary spend: inventory, contractors, marketing.

The bottom row. Beginning balance, plus receipts, minus disbursements, equals ending balance — which feeds the next column. Any week that goes below your minimum operating cushion gets highlighted. That highlight is the entire point of the exercise.

The Monday routine

A forecast built once is a souvenir. The value is in the cadence:

1. Every Monday, replace last week’s forecast column with actuals.
2. Note the misses — who did not pay, what cost more than planned.
3. Roll the model forward one week, so it always shows thirteen.
4. Read the bottom row. If a week has gone red, decide what moves this week.

The first week’s accuracy should be within a few percent almost immediately; week thirteen will always be fuzzy, and that is fine — by the time it matters it will be week four. The update takes about two hours once the model is set up, less if your books close cleanly and on time (a late close starves the forecast of good inputs — the case for fixing the close first is in the month-end close checklist).

Runway is the summary statistic

For funded companies, the forecast’s bottom row is also where honest runway math lives: current cash divided by forecast net burn, not last month’s burn, which annual prepayments can flatter badly (see deferred revenue for that trap).

More on forecasting and the reporting around it in Cash Flow & Forecasting. If you do not currently know your thirteen-week position, that is one of the eight questions in the books health check — and worth three minutes to find out what else you would flunk.