The 13-week cash forecast, built in an afternoon

A P&L tells you whether you’re profitable. A cash forecast tells you whether you can make payroll on the 15th. They’re different questions, and the second is the one that ends companies. A 13-week cash forecast is the first thing I build for every client. It takes an afternoon to set up and about 20 minutes a week to keep right.

Why 13 weeks

Thirteen weeks is one quarter. It’s short enough to forecast from real invoices, bills and payroll dates, and long enough to act on what you see. If the forecast shows a gap in week 10, you have two months to collect faster, delay a cost or arrange a credit line. If you find out in week 1, you have none of those options.

The layout

Columns are weeks, Monday to Sunday. Rows are:

• Opening cash, from the bank, not the books

• Receipts: customer collections, then any other income

• Payments: payroll and payroll tax, rent, suppliers, software, loan repayments, taxes and one-offs

• Net cash flow for the week

• Closing cash, which becomes next week’s opening cash

• Minimum cash, a line you never want to cross. Two months of payroll is a sensible start.

Build it in an afternoon

1. Start from today’s bank balances. Add up every account. The forecast is about cash you can spend, not what the accounts say you’ve earned.

2. List open invoices by due date. Then move each one to when the customer actually pays. If a customer always pays 18 days late, forecast it 18 days late.

3. Add new sales carefully. Include only signed work. For lots of small customers, use the weekly average of the last eight weeks.

4. Enter payroll exactly. Real dates and real amounts, including taxes and benefits. It’s your biggest payment and the easiest to get right.

5. Enter bills by due date. Start with your payables list, then add recurring costs from the last three months of bank statements.

6. Add the one-offs. Insurance renewals, annual software, tax payments, bonuses, a deposit on new equipment.

7. Check every week against your minimum. Any week below the line gets a note: what you’ll do about it, and by when.

Keep it right in 20 minutes a week

Every Monday, replace last week’s forecast with what actually happened. Write down the biggest difference and why it happened. Then roll the forecast forward one week, so there are always 13 ahead.

Track how accurate week one was. If it’s often off by more than 5%, one of your inputs is wrong, usually when customers pay. Fix the input, not the total.

What it shows you

A roaster I worked with was profitable every month and short of cash every month. The forecast showed why within a week. Green coffee was paid for months before it was roasted and sold, and wholesale customers took 52 days to pay. The fix was in the timing, not the margins. Over six months it freed $410,000.

That’s typical. A forecast rarely shows you something you didn’t half know. It shows you how big it is and when it lands, which is what you need to act.

One last rule

Arrange a credit line while the forecast still looks fine. Banks lend most easily to companies that don’t urgently need the money, and a well-kept forecast is the best evidence that you’ll pay it back.