The 13-week cash flow forecast, built in about 2 hours

A cash flow forecast has a reputation problem. It sounds like something an accountant builds for a bank, in a template with 40 rows, that gets opened twice and never again.

The version worth having is far smaller. It's a single sheet, 13 columns wide, that answers 1 question: in which week does the money run short. Building it is 1 of the money-block jobs in the guide to working on the business rather than in it, and it's the job that converts a vague worry into a date you can do something about.

The first build takes about 2 hours. Every update after that takes 20 minutes.

Why 13 weeks and not 12 months

13 weeks is a quarter, in the unit that cash actually moves in.

A monthly forecast hides the problem. A month that nets out positive can still contain a Thursday where wages, a supplier account and a finance payment all land in the same 3 days while your largest invoice sits unpaid at 48 days. Averaged across 4 weeks, that week disappears. Lived through, it's the whole event.

Go much beyond 13 weeks and the numbers stop being real. You're estimating jobs that haven't been quoted and payments from customers who haven't ordered. The value of this sheet is that almost everything on it is knowable: work already committed, invoices already issued, costs already contracted.

13 weeks is also long enough to act. A shortfall visible 8 weeks out can be solved by invoicing faster, collecting harder, moving a payment or bringing a job forward. The same shortfall visible on the morning it arrives can only be solved by borrowing or by not paying somebody.

What you need in front of you

Before you open a spreadsheet, gather 5 things.

1. Your current bank balance, as at the end of last week. Not the app's available balance, the actual cleared figure.
2. Your accounts receivable list, showing who owes what and how old each invoice is.
3. Your accounts payable list, showing what you owe and when it falls due.
4. Your fixed weekly and monthly outgoings. Wages and on-costs, rent, insurance, vehicle and equipment finance, subscriptions, utilities.
5. Your committed work. Jobs booked, quoted and accepted, with rough delivery dates.

Ask your accountant which regular obligations fall due inside the next quarter so nothing large surprises the sheet. That's a 10-minute conversation and it's worth having before the first build rather than after.

Build the sheet: 3 blocks of rows

Columns are the easy part: 13 of them, dated by the Monday of each week.

Rows split into 3 blocks.

Block 1, money in. Keep it in the categories that behave differently. Invoices already issued and awaiting payment. Work in progress you'll invoice this quarter. Deposits on jobs not yet started. Anything else, like an equipment sale, on its own line.

Block 2, money out. Wages and on-costs. Subcontractors. Materials and supplier accounts. Rent. Insurance. Vehicle and equipment finance. Loan repayments. Software and phones. Marketing. Owner drawings, which owners leave off constantly and which are a real outflow. Then a line for anything else your accountant flagged.

Block 3, the result. Opening balance for the week. Total receipts. Total payments. Net movement. Closing balance, which becomes next week's opening balance.

That's the entire structure. Resist adding rows for detail that doesn't change a decision. A forecast with 12 rows gets updated. A forecast with 60 gets abandoned in week 3.

Fill it in with what will happen, not what should

The build is mechanical. The honesty is the hard bit.

Put each receipt in the week the customer will actually pay, based on what that customer has done historically. If an account has paid at 55 days for 3 years, it pays at 55 days on your forecast, whatever your terms say. Owners who forecast to their stated terms build a sheet that's wrong by 4 weeks and then stop trusting it.

Place payments on their due date, not the date you'd prefer. Wage runs go in on the actual dates, including the quarters with a third run in a month.

For work not yet won, be conservative or leave it out. A forecast is not a sales target.

And keep the timing separate from profitability. A job can carry a strong margin and still land in the wrong week, which is the whole reason profitable businesses still run out of cash is a real phenomenon rather than a bookkeeping error.

Read it: the 3 questions it answers

An illustration with round numbers, not a real business. Opening balance $40,000.

All figures ex GST and illustrative only.

Read across and the sheet answers 3 questions immediately.

Where's the low point? Week 3, at minus $6,000. That business isn't unprofitable. Over the 6 weeks shown it takes in more than it pays out. It simply has 2 heavy payment weeks landing before 2 heavy receipt weeks.

How much buffer does it need? The distance between the opening balance and the lowest closing balance, plus a margin for the customer who pays late. In the illustration that's more than $46,000 of swing across 3 weeks, which tells the owner what a realistic minimum cash holding looks like.

Which lever fixes it? Week 3 is solvable several ways: invoice week 2's completed work immediately, collect the oldest debtor, take a deposit on a job starting in week 4, or move a discretionary payment by 10 days. All 4 are ordinary, and all 4 require knowing about week 3 while it's still 3 weeks away.

Update it every week, forever

The forecast's value is almost entirely in the update, not the build.

Same slot each week. Roll the window forward by 1 column so it's always 13 weeks. Replace last week's forecast figures with what actually happened. Then look at the variance and ask which assumption was wrong.

That variance is the most useful output of the whole exercise. It teaches you how your customers really behave, how long your jobs really take to invoice, and how optimistic you personally are, which turns out to be a stable and measurable trait.

Sit it alongside the quarter's plan so cash and priorities are read together rather than in separate conversations. How to write a 90-day plan you'll actually follow covers the other half of that pairing.

The mistakes that make a forecast useless

• Forecasting to your terms rather than to behaviour. The single most common error, and it makes every week optimistic.
• Leaving out owner drawings. The money still leaves the account.
• Building it monthly. It hides the exact week the trouble lands.
• Including work that hasn't been won. Turns the sheet into a wish.
• Building it once. A forecast that isn't updated weekly is a snapshot of a fortnight ago.

Your registered tax agent handles tax. A coach works on what the business does to make the money in the first place, which includes how fast the work is invoiced and collected. If the forecast shows the business can't meet what falls due, speak to your accountant and a qualified professional immediately rather than trading through it.

What to do next

Build the sheet this week, before the next quarter starts. Then put the update in the calendar as a repeating 20-minute appointment and treat it like a client booking.

There's no cash flow tool on this site. The calculators and quizzes here cover what coaching costs, whether you're at the point of wanting one, which kind of help fits your problem, and what an owner hour is worth in your business, so build the forecast in a spreadsheet or in your accounting file.

Then make sure the outputs land on the short list of figures you can quote without opening anything, which the numbers every owner should know sets out.