Many owners find out about a cash gap when payroll is two days away. A 13-week cash flow forecast shows it a month or more ahead, while you still have options. This guide covers the structure, how to forecast receipts from the way your customers actually pay, the weekly review that keeps it accurate and the mistakes that quietly break it.

Why 13 weeks?

Thirteen weeks is one quarter. That is long enough to see trouble coming (a slow-paying customer, a quarterly tax payment, a month with three payrolls) and short enough to forecast with real data instead of guesses. Most of the cash you will collect over the next three months is already sitting in your receivables. Beyond that, you are forecasting sales you haven’t made yet.

The forecast is weekly, not monthly, and that matters. A monthly view can show a healthy month-end balance while hiding a week in the middle where the account dips below zero. Payroll, rent and supplier runs land on specific days. A weekly view lines them up against the days customers really pay.

If your cash swings week to week, you’re far from alone. In the Federal Reserve’s Small Business Credit Survey, 51% of US employer firms cited uneven cash flow as a financial challenge, and the Fed notes this includes collecting receivables (2025 Report on Employer Firms).

The structure: five lines that matter

Every 13-week forecast has the same skeleton, with one column per week:

  1. Opening cash: the bank balance at the start of the week. For week 1, use the actual balance in the bank, not the balance in your books. After that, each week opens with the previous week’s closing cash.
  2. Receipts: cash you expect to collect, with your largest customers on their own lines and one line for everyone else, plus any other cash in, such as a loan draw or a tax refund.
  3. Disbursements: everything going out: payroll, rent, suppliers, taxes, debt payments, owner draws.
  4. Closing cash: opening cash plus receipts minus disbursements.
  5. Minimum cash line: the balance you never want to drop below. You set it, for example one full payroll plus rent plus a buffer for surprises.

Here are the first four weeks of an example forecast. The numbers are made up to show the mechanics.

LineWeek 1Week 2Week 3Week 4
Opening cash$84,000$63,500$72,000$35,000
Customer receipts$41,000$36,500$22,000$47,500
Payroll$38,000$0$38,000$0
Rent$9,500$0$0$0
Suppliers$14,000$18,000$21,000$12,500
Taxes$0$6,800$0$6,800
Debt payments$0$3,200$0$0
Closing cash$63,500$72,000$35,000$63,200
Minimum cash$45,000$45,000$45,000$45,000
Above (below) minimum$18,500$27,000($10,000)$18,200

Week 3 is the point of the exercise. Payroll and a big supplier run land in a light receipts week, and closing cash falls $10,000 below the minimum. Seen two weeks out, that is manageable: chase the specific invoices expected in weeks 2 and 3, agree a later date with a supplier, or arrange a draw on your line of credit. Seen on the Thursday before payroll, it is a crisis.

Forecasting receipts from real customer behavior

This is where forecasts usually go wrong. The easy approach assumes every invoice is paid on its due date. Customers don’t pay on terms. They pay on habits, and your forecast should use the habits.

Calculate each customer’s average days to pay

Export your paid invoices for the last six to twelve months with the invoice date and the date the money arrived. For each customer, count the days between the two and average them, weighted by invoice amount so one small invoice paid fast doesn’t skew the result. For customers with only one or two invoices, use their terms plus your overall average delay.

Worked example

Say three invoices go out in week 1 of the forecast. Here is where each lands based on terms, and based on history.

CustomerTermsAverage days to payInvoiceWeek by termsWeek by history
Customer ANet-3033$18,000Week 5Week 5
Customer BNet-3047$12,400Week 5Week 7
Customer CNet-4571$9,600Week 7Week 11

By terms, you expect all $40,000 by the end of week 7. By history, you expect $30,400, and the last $9,600 doesn’t show up until week 11. If week 7 carries a payroll, that $9,600 is the difference between a calm week and a call to the bank.

What to keep out of receipts

  • Disputed invoices: put them on a separate “at risk” line below the forecast until the dispute is resolved.
  • Invoices over 90 days: forecast zero unless the customer has committed to a date.
  • Retainage: on construction jobs, forecast it when the release conditions are met, not with the progress payment.
  • Invoices not sent yet: include expected billings such as retainers or progress billings, but apply the same customer payment pattern.

Where AR follow-up comes in

History tells you how a customer usually pays. Follow-up tells you what is happening this week. When someone confirms receipt of an invoice before the due date, calls AP when it goes late and gets a specific payment date, that invoice moves from “projected” (based on history) to “committed” (the customer gave a date and an amount). Track the two separately. A forecast where most of the next four weeks is committed is one you can make decisions on.

Follow-up also surfaces disputes early, so they move to the at-risk line before they blow a hole in a week you were counting on. And consistent follow-up shortens each customer’s average days to pay, which pulls cash forward in the forecast. If nobody on your team has time for it, that is the gap an invoice follow-up service fills.

Forecasting disbursements

Outflows are easier to predict because you control most of them. The risk is forgetting the ones that don’t come every week.

  • Payroll: on actual pay dates, including employer taxes and benefits. If you pay every two weeks, mark the months with three pay dates.
  • Rent and leases: fixed amounts on fixed dates, including equipment and vehicle leases.
  • Suppliers: start from your accounts payable aging, then add expected purchases. Use your real payment habits: if you pay suppliers on day 40, forecast day 40.
  • Taxes: sales tax or GST/HST, payroll remittances, estimated income tax. Get the dates and amounts from your accountant.
  • Debt: loan payments, interest on your line of credit, credit cards you pay in full.
  • The irregular items: annual insurance premiums, software renewals, owner draws, planned equipment purchases. These wreck forecasts precisely because they show up once a year.

Leave out anything that doesn’t move cash, like depreciation. A cash forecast is not a profit and loss statement.

The weekly review: variance and rolling forward

A forecast built once and never touched is a guess in a nice format. The value comes from a short review every week.

  1. Enter last week’s actuals next to what you forecast, line by line.
  2. Explain the big variances. Sort each one into timing (the money is coming, just later) or permanent (it isn’t coming, or it will be smaller).
  3. Update the remaining weeks with what you learned: new payment promises, new disputes, changed supplier dates.
  4. Roll forward. Drop the week that just ended and add a new week 13, so you always see a full quarter ahead.

Example: you forecast $41,000 of receipts last week and $33,500 arrived. Of the $7,500 gap, $6,000 is Customer B paying a week later than usual (timing, so it moves to this week) and $1,500 is a short payment over a pricing disagreement (permanent until resolved, so it moves to the at-risk line). Two notes, and this week’s forecast is better than last week’s.

Keep a simple log of variances. If receipts come in below forecast week after week, your averages are too optimistic. Adjust them instead of hoping it evens out.

Common mistakes that break a cash forecast

  • Forecasting receipts on payment terms. Use each customer’s real average, as shown above.
  • Starting from the book balance. Uncleared checks and pending deposits make it different from the cash you can spend.
  • Counting disputed or very old invoices as cash. They belong on the at-risk line.
  • One lump for all receipts. Forecast your largest customers individually. A single total hides which one is late.
  • Forgetting irregular payments. Quarterly taxes, annual premiums, the third payroll in a month.
  • No minimum cash line. Without it, every balance above zero looks fine.
  • Building it and walking away. If it isn’t rolled forward weekly, it is out of date within a month.

You don’t need special software to avoid any of this. A spreadsheet with one column per week works for most small businesses, fed by exports of open invoices and bills from QuickBooks Online, Xero or your ERP. The free toolkit includes a 13-week cash forecast template set up this way, plus an AR aging template that feeds the receipts side.

Want the receipts side built for you?

Extra Myle builds a rolling 13-week forecast from your open invoices and each customer’s real payment behavior, then updates it weekly with the promises and disputes we hear during follow-up.

Book a free AR review

To see what that weekly package includes, read about our AR reporting and cash flow forecasting service.

FAQ

What is a 13-week cash flow forecast?

It is a week-by-week projection of the cash coming into and going out of your bank account over the next quarter. Each week shows opening cash, receipts, disbursements and closing cash, so you can see a shortfall before it happens and act early.

How is it different from a budget?

A budget plans revenue and expenses, usually by month, for a year. A 13-week forecast tracks actual cash timing: when customers really pay and when bills really go out. You can be on budget and still run short of cash in a given week.

How accurate should the forecast be?

The next two or three weeks should be close, because most of those receipts are already invoiced and many are committed. Later weeks are rougher. Accuracy improves when you review variances every week and base receipts on how each customer actually pays.

Who should own the forecast?

Usually the owner, a bookkeeper or a fractional CFO. Whoever owns it needs fresh AR input each week, including payment promises and disputes, or the receipts side drifts away from reality.

EM
Extra Myle team

We run invoice follow-up, billing disputes and weekly AR reporting for B2B companies in the United States and Canada, in our clients' names. About us.