The 13-Week Cash Flow Forecast: A Complete Guide for CFOs

Every Monday morning, a CFO somewhere opens Chase, BofA, and their payroll bank, copies the balances into Excel, and emails a "cash update" to their CEO. An hour of their life, gone. The document they're trying to produce is a 13-week cash flow forecast. It's the single most important operating document a CFO owns — and most finance teams build it the hard way.

This guide explains what a 13-week cash forecast is, why 13 weeks is the magic number, how to build one, the mistakes every finance leader makes the first time, and when to stop building it by hand. If you want to skip ahead, we've made our entire model free as an Excel template — same structure our paying customers use.

What a 13-week cash flow forecast actually is

A 13-week cash flow forecast (sometimes called a "TWCFF" or a "direct method cash forecast") is a week-by-week projection of your cash position over the next 13 weeks — one full quarter — broken into three sections:

1. Opening cash — total cash across every operating account
2. Cash inflows — customer collections, financing draws, one-offs
3. Cash outflows — payroll, vendor AP, rent, SaaS, debt service, taxes
4. Closing cash — the number that rolls forward to next week

The output is a table that looks identical to what banks and lenders expect to see when they ask "what's your forecast?" There's no alternative format that's more standard.

Why 13 weeks? Why weekly?

The "13 weeks" number isn't arbitrary. It's the shortest forward-looking horizon that captures:

• A full quarter of operational cadence (AR cycles, payroll runs, quarterly taxes)
• Enough time to see a crisis coming and respond before you run out of cash
• Short enough that the forecast is actually accurate

Monthly forecasts are too coarse — you miss within-month cash crunches (payroll on the 15th while major AR clears on the 18th). Daily forecasts are too granular; they create noise that masks signal. Weekly is the Goldilocks zone. Every mid-market CFO I know uses 13 weeks, weekly.

Why it matters more than anything else you build

Your P&L lies. Your balance sheet lies. Your cash forecast doesn't.

The P&L says you're profitable, but you booked a $500K receivable that won't clear for 90 days. Payroll is in 14 days. The cash forecast is the only document that tells you the truth: you're going to be short.

Every catastrophic small business failure I've watched up close was a surprise to the CEO — but visible in the cash forecast six weeks earlier, if anyone had been looking.

How to build one: the step-by-step

1. Pull opening balances

Log into every operating account. Grab today's balance. Sum them. That's your starting cash position.

Shortcut: if you use a multi-bank aggregator, this is one number already sitting on a dashboard.

2. Project inflows

For each of the next 13 weeks, estimate:

• AR collections based on invoice aging. If you have $420K aged 0–30 days, assume roughly 70% clears in the next 14 days (tune this to your actual customer behavior)
• One-offs — loan draws, tax refunds, grants, deposits from known contracts
• Other — miscellaneous revenue, interest income, etc.

Don't guess. Use your aged AR report and your actual customer payment history.

3. Project outflows

This is where most forecasts go wrong. Be ruthlessly specific:

• Payroll — dates are known, amounts are known. Don't approximate
• AP — vendor invoices you've received, dated by due date
• Rent/lease — fixed, predictable
• SaaS subscriptions — go through your credit card statement once and map every recurring charge to its exact week
• Debt service — principal plus interest, by lender, by date
• Quarterly taxes — put them in the correct week, not the correct month

4. Calculate closing cash each week

Closing cash = opening + inflows − outflows. Next week's opening = this week's closing. Roll the chain forward 13 weeks.

5. Add a minimum cash floor

Set a number you'd be nervous going below. Most CFOs pick 30 days of payroll. When the closing balance in any week drops below the floor, turn that cell red. Now you have an alert you'll actually notice.

6. Track variance versus prior forecast

Every Monday, before you update, copy last week's closing balances into a "prior forecast" row. After you update, a variance column shows how much each week moved. If things are moving more than 10%, your forecast inputs are stale — go investigate.

The mistakes every CFO makes the first time

1. Using monthly buckets. Disguises payroll timing problems. Use weekly.
2. Forgetting semi-monthly payroll is different from biweekly. The 15th and the end of the month are not "every two weeks."
3. Optimistic AR collection assumptions. If your 31–60 aging bucket is 40% of AR, your cycle is longer than you think it is.
4. Missing quarterly items. Tax payments, insurance renewals, bonuses, annual SaaS invoices.
5. Not updating every Monday. A forecast more than a week stale is misleading. Update on a cadence or throw the model away.
6. Treating it as a planning document, not an operating one. The forecast isn't what you want to be true — it's what's going to happen. Lean into the bad news.

When to stop building it by hand

If you're updating your forecast once a week and the update takes you under 30 minutes, keep going. You're fine.

If the update takes you 2+ hours, or if the numbers are out of date by the time you present them to the board, that's the signal to automate.

Automated tools pull bank balances via Plaid, consolidate across accounts, and roll the forecast forward every night. You spend Mondays interpreting the forecast, not rebuilding it. Most CFOs who switch say the hardest part was trusting the automation — once they did, they never went back.

A working definition of "good"

A 13-week cash forecast is working well if:

• You update it in under 30 minutes a week
• Your variance vs the prior week is under ±5% on the closing cash row
• Your minimum cash floor is set and you'd trust the red cells
• Your CEO can read it in 15 seconds and know if they should be worried
• Your lender would accept it without questions

If any of those are false, the forecast is decorative, not operational. Fix those before you do anything else.

Frequently asked questions