How to build a 13-week cash flow forecast and use it every Monday

Your monthly forecast says you'll end the quarter with $74,000. That's true. It's also true that there's a week in the middle where you'd have to choose between payroll and the tax bill, and the monthly view never shows it.

That's the gap a 13-week cash flow forecast closes. It's the tool lenders, boards and turnaround advisers ask for first, and it's one of the simplest things in finance to build. This guide walks through one company's quarter.

The quick answer

A 13-week cash flow forecast is a week-by-week, cash-only view of the next quarter: what lands in the bank, what leaves it and what's left at the end of each week. You update it every Monday, compare last week's forecast with what actually happened and add a new week at the end, so you always see 13 weeks ahead.

Why weekly, why 13

Weekly because cash problems happen in weeks, not months. Payroll every two weeks, a quarterly tax payment, an annual renewal and one late customer can all land in the same seven days.

Thirteen because it's a quarter: far enough ahead to act (delay a hire, chase an invoice, open a credit line), near enough that the numbers are mostly known rather than guessed.

Cash-only because it's not a P&L. No depreciation, no accruals, no revenue recognition. Money in, money out, on the day it actually moves.

A worked example

A company that raised last year has:

• $226k in the bank and a $150k minimum buffer it won't go below
• About $11k of customer receipts a week, growing 1% a week
• $6.5k a week of other payments (suppliers, software, rent)
• $34k of payroll every two weeks
• Three known one-offs: an $18k annual software renewal in week 3, a $24k customer invoice landing in week 5, a $21k quarterly tax payment in week 9

Here's how the quarter plays out:

What the table tells the founders that the monthly view didn't:

• Cash first drops below the $150k buffer in week 8, a payroll week.
• It stays below the buffer for 6 of the 13 weeks.
• The low point is $68.5k in week 12, again a payroll week.
• Average burn is about $11.7k a week, so the $74k left at week 13 is roughly six more weeks of cash. That's the number that should start a conversation today, not in week 12.

Model the fix before you need it

The value of a 13-week forecast is that you see the problem with weeks to spare. Here's what three changes do to the same quarter:

None of these alone keeps the company above its buffer. Together they buy enough time to raise, to arrange a credit line or to make a bigger decision calmly. That's the purpose: turn a surprise into a plan.

How to build yours in an afternoon

1. Start with today's bank balance. Not the accounting balance; what the bank says.
2. List receipts by week. Use your accounts receivable aging: who owes what, and when they've actually paid in the past.
3. List payments by week. Payroll dates, rent, loan repayments, supplier runs, card bills, taxes.
4. Add the one-offs you already know about. Renewals, deposits, equipment, legal fees.
5. Set a buffer. The lowest balance you'd be comfortable holding, often a month or two of payroll.
6. Flag every week below it. Those are the weeks that need a plan.

The Monday routine

A 13-week forecast is only useful if it's current. Each Monday:

1. Enter last week's actuals next to last week's forecast.
2. Look at the variance. Were receipts late? Did a payment come early? Variances that repeat are a forecasting problem; fix the assumption.
3. Drop the old week, add a new week 13.
4. Check the low point and the buffer weeks. Did either move?
5. Send a two-line update to your co-founder, or your board if cash is tight.

Twenty minutes a week. It's the cheapest insurance a founder can buy.

How it fits with your other forecasts

The 13-week view sits underneath your monthly model, not instead of it. The monthly model answers "how long is our runway and how much do we need to raise?" (see startup runway: how much is enough). The 13-week view answers "can we make every payment this quarter, and which week is tight?" Your board wants both, alongside budget vs actuals; the variance calculator covers that third view.

To see the routine in practice, the case studies include an AI company whose Monday 13-week forecast absorbed a 19% cost overrun after a large customer launched early.