Cash flow
How to build a 13-week cash flow forecast
November 17, 2026 · 10 min read
Quick answer
A 13-week cash flow forecast lists, week by week, the cash you expect to collect and the cash you expect to pay out, starting from your actual bank balance today. Each week's ending cash becomes the next week's opening cash. Build it from collections and payment dates (not revenue and expenses), set a minimum cash floor, and roll it forward every week so you always see three months ahead.
Most small businesses don't run out of money in a bad year. They run out in a bad week. Payroll lands on the same Friday as the insurance renewal, your biggest client pays ten days late, and suddenly an otherwise healthy company is calling the bank.
A monthly budget won't catch that, because it averages the month out. A 13-week cash flow forecast will. It's one quarter, broken into weeks, and it's the single most useful spreadsheet a growing business can keep. It's also simpler than people expect.
This post shows you how to build one, what goes in it, a worked example with a real-looking cash dip, and the 30-minute weekly routine that keeps it honest.
13 weeks
One quarter, long enough to see payroll cycles, tax payments and loan payments coming
Weekly
How often you update it. A cash forecast updated monthly is already out of date
1 number
The one that matters most: the lowest ending cash in the next 13 weeks
Why 13 weeks and not a monthly budget?
Thirteen weeks is a quarter. It's short enough that you can predict most of it with real information (open invoices, scheduled payroll, known bills) and long enough to see the problems while you can still do something about them.
The weekly view is the key part. Cash moves in lumps. Payroll might go out every other Friday, rent on the 1st, a loan payment on the 15th, a quarterly tax payment once every few months. A monthly budget that shows "June: +$12,000" can hide a week in the middle where you're $20,000 short. Weekly buckets show that week.
It's also different from a profit forecast. A budget asks, "Will we make money this quarter?" A 13-week forecast asks, "Will we have enough cash in the bank every single week?" You need both, but the second one is what keeps payroll from bouncing. (If profit and cash disagreeing confuses you, start with our post on cash flow vs profit.)
What goes into a 13-week cash flow forecast?
Keep it to one sheet. Columns are weeks 1 to 13. Rows are:
- Opening cash. Week 1 starts with your actual, reconciled bank balance. Every later week starts with the previous week's ending cash.
- Cash in. Customer collections, broken out for your biggest customers. Any loan draws, owner contributions or other one-offs, on their own lines.
- Cash out. Payroll (including taxes and benefits), rent, loan payments, regular vendor bills, software and subscriptions, taxes, and any one-off payments.
- Net cash flow. Cash in minus cash out.
- Ending cash. Opening cash plus net cash flow.
- Cash floor. The minimum you're willing to hold. Any week below it is a flag.
That's genuinely it. You can add detail later, but a forecast with 40 rows that nobody updates is worse than one with 12 rows that's current every Monday.
Starting point
- Today's bank balance, reconciled (not the balance in your accounting software if the feed is behind)
- Any money sitting in payment processors like Stripe, Square or PayPal
Cash coming in
- Accounts receivable aging, by customer
- How many days each major customer actually takes to pay (not your stated terms)
- Signed work you'll invoice in the next 13 weeks, and when you'll invoice it
- Any loan proceeds, grants or owner money expected
Cash going out
- Payroll calendar with pay dates and total cost per run
- Accounts payable aging and vendor payment terms
- Rent, loan and lease payment schedules
- Tax payments: payroll tax deposits, sales tax, estimated income tax
- Annual or one-off bills: insurance renewals, software contracts, equipment
How to build it, step by step
Step 1: Start from a real number. Use the actual bank balance as of today, after you've checked it against the bank. If your books are three months behind, fix that first, because every week in the forecast inherits the error.
Step 2: Forecast collections, not sales. This is where most forecasts go wrong. If you invoice $25,000 this week and your customers pay in 40 days, that cash shows up in week 6 or 7, not week 1. Use each big customer's real payment habit. For smaller customers, use an average.
Step 3: Put every payment in the week it leaves the bank. Payroll goes in the week of the pay date. A bill due on the 20th with 30-day terms goes in the week you'll actually pay it. Don't spread annual costs evenly. The point is to see the lumps.
Step 4: Don't forget taxes and debt. These are the two items forecasts miss most. If you run payroll yourself, IRS Publication 15 sets your deposit schedule: if you reported $50,000 or less of employment tax in the lookback period, you deposit monthly, with taxes on a month's payments due by the 15th of the following month. Above $50,000, you deposit semiweekly, within days of each payday. Owners of pass-through businesses often pay quarterly estimated tax too; the IRS due dates for individuals are April 15, June 15, September 15 and January 15. Loan payments go in at the full amount (principal and interest), because that's what leaves the bank.
Step 5: Set your cash floor. A simple starting point is enough to cover one payroll plus a week of regular bills. If any week dips below it, you have a decision to make now, not then.
Step 6: Roll it every week. Each Monday, replace last week's forecast with what actually happened, drop it off the left side, and add a new week 13 on the right. That's what makes it a rolling forecast.
A worked example: spotting the dip in week 7
Say you run a 10-person IT support company. You start the quarter with $52,000 in the bank. Payroll costs about $24,000 every other week, rent is $7,500 a month, your equipment loan payment is $3,000 a month, and regular vendors and software run about $5,000 a week. You set a cash floor of $30,000: one payroll plus a week of bills, rounded up.
Collections vary between about $16,000 and $30,000 a week. Two things land in week 7: your annual business insurance renewal ($14,000) and your quarterly estimated tax payment ($12,000). And week 7 collections are light, because your largest client tends to pay late in that part of the cycle.
Here are weeks 5 to 8 of the forecast:
| Week 5 | Week 6 | Week 7 | Week 8 | |
|---|---|---|---|---|
| Opening cash | $64,500 | $49,000 | $66,000 | $27,000 |
| Customer collections | $21,000 | $25,000 | $16,000 | $23,000 |
| Payroll | −$24,000 | −$24,000 | ||
| Rent | −$7,500 | |||
| Loan payment | −$3,000 | |||
| Vendors and software | −$5,000 | −$5,000 | −$5,000 | −$5,000 |
| Insurance renewal and estimated tax | −$26,000 | |||
| Ending cash | $49,000 | $66,000 | $27,000 | $45,000 |
And here's the full 13 weeks:
Projected ending cash, next 13 weeks
Hypothetical 10-person IT support company
Hypothetical example. Weeks 5 to 8 match the table above.
Show the numbers
| Ending cash | |
|---|---|
| W1 | $37,500 |
| W2 | $55,500 |
| W3 | $45,500 |
| W4 | $64,500 |
| W5 | $49,000 |
| W6 | $66,000 |
| W7 | $27,000 |
| W8 | $45,000 |
| W9 | $38,500 |
| W10 | $57,500 |
| W11 | $51,500 |
| W12 | $75,500 |
| W13 | $71,500 |
Look at the shape. The business is fine. Cash ends the quarter at $71,500, higher than it started. But week 7 dips to $27,000, $3,000 under the floor, with the next payroll due in week 9. A monthly budget would have shown a perfectly healthy month and nothing else.
Because you see it six weeks out, you have easy options:
- Ask the insurer about paying the renewal in installments.
- Call your largest client in week 5 and confirm when the payment will go out.
- Hold off on a planned equipment purchase or owner draw until week 8.
- Line up a credit line draw now, so it's there if you need it.
Any one of those fixes the dip. Found the Thursday before, you'd have far fewer choices.
The weekly update routine
The forecast is only useful if it's current. Here's a routine that takes about half an hour once it's set up.
Step 1
Reconcile the bank
Confirm last week's actual ending cash against the bank, not the forecast.
Step 2
Actual vs forecast
Replace last week's forecast with actuals and note anything off by more than about 10%.
Step 3
Update collections
Check the receivables aging. Move any late invoice to the week you now expect it.
Step 4
Update payments
Add new bills, confirm payroll, and check for anything annual coming up.
Step 5
Roll forward
Drop the finished week and add a new week 13.
Step 6
Check the floor
Find the lowest week. If it's under the floor, decide what to do now.
Step 2 matters more than it looks. When you compare the forecast to actuals every week, you find out which assumptions are wrong (usually customer payment timing) and the forecast gets sharper. After a couple of months, you'll probably trust it more than your gut.
Common mistakes
Using revenue instead of collections. Revenue is when you earn it. Cash is when they pay. Mixing them up makes the forecast look rosier than reality, every time.
Starting from an unreconciled balance. If the accounting software is missing two weeks of transactions, so is your forecast. Start from the bank.
Leaving out taxes, debt and annual bills. These are exactly the lumps the forecast exists to catch. Payroll tax deposits, sales tax, estimated tax, loan payments, insurance renewals: get them all on the calendar.
Building it once and never touching it again. A static forecast is a guess with a date on it. Rolling it weekly is what makes it useful.
Overbuilding. You don't need a 40-tab model. If updating it feels like a project, you'll stop doing it.
When to get help
If you have a handful of customers, one bank account and a predictable payroll, you can build and run this in a spreadsheet yourself. Honestly, every owner should try it once, because it teaches you where your cash really goes.
It's worth getting help when:
- your books aren't current, so there's no reliable starting balance
- you have lumpy, project-based income or lots of customers on different terms
- cash is already tight and you need to make decisions about payroll, hiring or borrowing
- you're preparing to talk to a lender, who may well ask for a 13-week forecast
Building and running a weekly cash forecast is part of what a fractional CFO does, and it works best on top of books that close every month. If you'd like to talk through your own situation, book a free call.
Frequently asked questions
What's the difference between a 13-week cash flow forecast and a budget?
A budget plans profit, usually by month, for a year. A 13-week forecast predicts actual cash in and out, by week, for the next quarter. The budget tells you whether the plan makes money; the forecast tells you whether you can pay what's due every week along the way.
Can I build a 13-week cash flow forecast in Excel or Google Sheets?
Yes, and for most small businesses a spreadsheet is the right tool. One row per type of cash in or out, one column per week, and a formula that carries ending cash into the next week's opening cash.
How accurate should it be?
The first few weeks should be close, because you already know most of what's coming. Weeks 10 to 13 will be rougher. That's fine. Comparing forecast to actual every week is what tightens it up over time.
What should my cash floor be?
It depends on how lumpy your costs are and how predictable your collections are. A reasonable starting point is one payroll plus a week of regular bills. Businesses with seasonal income or slow-paying customers often want more.
Who needs a 13-week forecast?
Any business where a single bad week would hurt: growing companies, project-based businesses, anyone with tight cash or debt covenants, and anyone talking to a lender. Lenders sometimes ask for one too.
The bottom line
A 13-week cash flow forecast is a quarter broken into weeks, built from when cash really moves. Start from a reconciled bank balance, forecast collections and payment dates, include taxes and debt, and set a cash floor. Then roll it every Monday. The payoff is simple: you see the bad week six weeks early, while fixing it is still a phone call and not a crisis.
Sources
This article is general information, not advice for your situation. Rules change and details matter, so talk to us or your CPA before acting on it.
More articles
- Cash flow vs profit: why your P&L and bank balance disagree
November 17, 2026
- 1099-NEC vs 1099-MISC: which to file, deadlines and how-to
November 17, 2026
- ASC 606 in plain English: the five steps every owner should know
November 17, 2026
Have a question about this article?
Ask us. A real person will reply by email within one business day. Your question isn’t published.
Ready for books you can trust?
Book a free 30-minute consultation. We'll review where your finances stand today and show you exactly how we'd help.