How to Build a 13-Week Cash Flow Forecast, Line by Line

To build a 13-week cash flow forecast, start with your reconciled bank balance, lay out thirteen weekly columns, and fill in cash coming in by source and cash going out by category, week by week, from documents you already have: open invoices, bills due, the payroll calendar, loan schedules, and the estimated tax dates. The bottom row is your projected bank balance each Friday. It takes about an hour to build the first time and twenty minutes a week to keep current.

We run an accounting firm, and we wrote earlier about why this forecast is the one report every owner should see weekly. This is the companion: how it’s actually built, row by row, so you can do it or know what you’re paying for.

Start with a number you trust

The first cell is the bank balance today, and it has to be the reconciled balance, not the online banking number. The online number ignores checks you’ve written that haven’t cleared and deposits in transit. The reconciled book balance accounts for both. If the books aren’t reconciled through last month, the forecast is wrong from cell one, and every week after inherits the error.

If you have multiple operating accounts, forecast the total, but keep the reserve or tax savings account out unless you plan to draw on it. The forecast is about the cash you’ll actually use.

The columns

Thirteen weeks, each column ending on Friday. Why thirteen: it’s one quarter, long enough to see the next estimated tax payment and two or three rent cycles, short enough that the numbers are real rather than budgeted. Week one is this week. Week thirteen is a quarter out.

Cash in, by source

Don’t forecast “revenue.” Forecast where money actually arrives, because each source has its own timing.

• Open invoices. Pull the receivables list. Put each invoice in the week you expect it to be paid, not the week it’s due. If a customer’s terms say net 30 and their history says 52 days, use 52. Your books know this; the aged receivables report and last year’s payment history tell you which customers are net 30 and which are net whenever.
• Recurring or contract revenue. Retainers, maintenance contracts, rent. Same amount, same week each month.
• Card and point-of-sale deposits. For a restaurant or retail shop, use last year’s same weeks, adjusted for what you know. Processor deposits lag sales by a day or two; put them in the week they land.
• Deposits on new work. Only the ones that are signed. A proposal isn’t cash.
• Anything else: a loan draw you’ve already arranged, a tax refund with a date, an owner contribution you’ve decided on.

The discipline is that every dollar in the “in” rows comes from a document or a pattern, not a hope.

Cash out, by category

Out is easier, because most of it is scheduled.

• Payroll. Every pay date, gross wages plus employer taxes plus any 401(k) match, on the date it hits the bank. If you pay biweekly, some months have three payrolls; the forecast is how you see those coming.
• Payroll tax deposits. On their schedule, separate from payroll if they land on a different day.
• Rent and fixed occupancy. Rent, utilities, insurance, on their dates.
• Loan payments. Every loan and line, full payment amount (principal and interest both leave the bank, even though only interest is an expense), on the due date.
• Vendor bills. From the payables list, in the week you’ll pay them, which for most businesses is the week they’re due. Add the recurring ones that don’t get entered as bills: software, phone, subscriptions.
• Cost of sales that isn’t on a bill yet. For a restaurant, next week’s food purchases; for a contractor, materials on jobs already scheduled. Use the run rate from closed months.
• Estimated tax payments. The fourth-quarter individual estimate is due January 15, 2027. C corporations’ fourth installment is December 15, 2026. Put them in. They’re the ones that get forgotten and the ones that land hardest.
• Owner draws. What you actually take, when you take it.
• Known one-offs. The equipment purchase, the insurance renewal, the annual software bill.

The bottom row

Beginning balance, plus in, minus out, equals ending balance, which becomes next week’s beginning balance. Thirteen times. That row is the forecast.

Read it for two things. The lowest point: which week, and how low. And the direction: is the balance thirteen weeks out higher or lower than today? A low point in week six that recovers by week nine is a timing problem with a timing solution. A balance that drifts down every week is a pricing or cost problem, and no forecast fixes that; it just tells you sooner.

How do I know if I can make payroll next month?

Find the payroll dates in the “out” rows and look at the ending balance the Friday before each one. If it’s positive by a comfortable margin, yes. If it’s thin, the forecast tells you which invoices need to arrive first, and you can call those customers this week instead of the morning payroll is due. If it’s negative, you have weeks of notice, which is enough to draw on a line, delay a purchase, or move a vendor payment. The forecast doesn’t create cash. It creates time.

The weekly update

Twenty minutes, every Monday. Replace the week that just ended with actuals from the reconciled books. Push the invoices that didn’t get paid forward to the week you now expect. Add the new invoices and bills from last week. Add week fourteen at the end. Look at the low point again.

The update is where the forecast earns its keep. A forecast built once and left alone is a budget. A forecast updated weekly is a picture of the next quarter that’s never more than a week stale, and the act of pushing a customer’s invoice forward for the third Monday in a row is how you learn that “net 30” was never true.

What a CFO does with this number

A fractional CFO reads the forecast every week and turns the low point into a decision, before it arrives. Sometimes that’s a collections call on two specific invoices. Sometimes it’s timing a vendor payment. Sometimes it’s the conversation about a line of credit, sized to the deepest negative week plus a margin, set up while the balance is healthy so the bank says yes.

The CFO also uses the forecast to answer the questions owners actually ask. Can I hire in November? Can I take a bigger draw this quarter? Should I buy the truck in December for the depreciation, or wait? Every one of those is a “what if” on the forecast: add the row, look at the bottom line, decide. Without the forecast those questions get answered by looking at today’s balance, which is the worst possible input.

Where we come in

The forecast only works on top of closed books: a reconciled starting balance, a real receivables list, loans that match the lender. Our monthly close produces all of that, and for advisory clients we build the 13-week forecast from it and update it weekly, with a short note when the low point moves. In-house payroll means the payroll dates and amounts in the forecast are the real ones, not an estimate of what a third party will pull.

If you check the banking app every Monday and hope, reach out. We’ll start with a two-hour diagnostic, get the books to a reconciled balance, and build the forecast from there.

Frequently asked questions

What is a 13-week cash flow forecast?
A weekly grid, thirteen weeks out, starting from the reconciled bank balance and walking forward with cash in by source and cash out by category.

What does it catch?
The negative week where two payrolls and a quarterly payment land together, and the customer whose net 30 is really net 52.

How long does it take?
About an hour to build from closed books and twenty minutes a week to keep current.