The 13-Week Cash Forecast: How to Build One on Your Own Numbers

For three weeks we’ve been building tools for the same hour. The Alignment Snapshot to agree on what matters. The Priority Scoreboard to put one owner on every line. The 60-Minute Agenda to turn “at risk” into a decision.

Every one of them assumes something: that the business will still have the cash to act on what it decides.

This week we build the tool that tells you whether it does.

Back in Issue 4 I laid out why profitable businesses still run out of money. Profit is measured over a period. Cash is what’s in the account on a Tuesday. I also promised a simple way to build a forward cash view on your own numbers. Here it is.

Why Thirteen Weeks

Thirteen weeks is one quarter. That length isn’t tradition. It’s arithmetic on two competing needs.

Short enough to be specific. You know roughly who owes you what, and when payroll lands. Push out to twelve months and you’re not forecasting. You’re guessing with a spreadsheet.

Long enough to see a problem while it’s still solvable. A gap you spot in week nine has eight weeks of decisions in front of it: collections calls, a deposit on a big job, a conversation with the bank while you still look like a good risk. The same gap found the Thursday before payroll has none of that. It has one option, and it’s expensive.

That’s the whole value. Not prediction. Lead time.

Start From the Bank, Not the P&L

Here’s the discipline it rests on, and the one most owners get wrong the first time.

A cash forecast records when money moves, not when revenue is earned.

Your profit and loss statement (the P&L) books the sale in the month you did the work. Your forecast books the cash in the week the customer’s payment clears. For a business paid in 45 days, that’s a month and a half between the two. An invoice sent late in a quarter collects its cash in the next one. Your accounting system is not wrong. It’s answering a different question. Do not start from it.

Start from the bank balance. Today’s actual number.

The Tool

Open a spreadsheet. Thirteen columns, one per week, dated by the week ending. Then six rows.

1. Starting cash. Week one is your real bank balance today. Every week after is the prior week’s ending cash. This is the only link between columns.

2. Cash in, customer receipts. Not sales. Collections. Take your receivables aging (who owes you, and how long it’s been outstanding) and place each invoice in the week you actually expect the money. Use observed behavior, not stated terms: a customer on 30-day terms who has paid at 45 for three years goes in week seven, not week five. That one honesty adjustment is worth more than any formula in the sheet. Add new work only where it’s genuinely likely: signed, deposited, or on a standing contract.

3. Cash in, everything else. Loan draws, tax refunds, an owner contribution, proceeds from selling a truck. Short list, real money.

4. Cash out, the fixed and the known. Payroll on its actual pay dates, with the payroll taxes that ride along. Rent. Loan and lease payments. Insurance premiums you pay monthly. Subscriptions. Your quarterly estimated tax payment, the one your accountant has already told you the date and amount for. This is the bulk of what leaves the account, and you already know nearly all of it. Fill them in first; it’s the fastest part of the build.

5. Cash out, the variable. Supplier and subcontractor payments (your payables, the bills you owe), placed in the week you’ll actually pay them. Materials for scheduled work. Owner draws: real money out, and leaving them off is how a forecast quietly lies to you. Then the lumpy one-offs owners forget: equipment, the annual insurance renewal, a deposit on a new location.

6. Ending cash. Starting, plus in, minus out. Carry it into the next column.

Six rows, thirteen columns, no formula harder than addition.

The Number That Matters Is the Low Point

Most people build this, look at week thirteen, see a healthy balance and relax. That’s reading it wrong.

The number that matters is the lowest ending balance anywhere in the thirteen weeks, the trough, and which week it lands in. A quarter that ends at $300,000 and dips to $9,000 in week nine is not a comfortable quarter. It’s a quarter with one bad week in it, and you now know its date.

So draw a second line: your floor. The balance below which you genuinely cannot operate, usually a payroll run plus a month of fixed costs. Every business has this number. In three decades I’ve rarely seen one written down.

Now the forecast has one job: tell you, as early as possible, when the trough is heading for the floor.

Here’s the shape it takes. It’s a composite, drawn from a pattern I’ve watched repeatedly rather than any one company. A services business, comfortably profitable, no concerns at all. Building the forecast, the owner finds a quarterly tax payment, an annual insurance renewal, and a three-week gap in collections all landing in the same seven days. No single item is a problem. Together, in that one week, they are. Nothing was wrong with the business. The calendar was stacked, and nobody had put the stack on one page.

Found in week two, that’s a scheduling conversation. Found in week nine, it’s a loan application at the worst possible moment.

Twenty Minutes, Every Monday

A forecast built once is a document. Built weekly, it’s a rhythm. Only the rhythm works.

Every Monday, one owner does three things, in this order. Compare last week’s forecast to what really happened. Do this first, before anything gets overwritten. Update the bank balance to today’s actual. Roll it forward: drop the finished week and add a new week thirteen, so the horizon stays a full quarter out.

That first one is the whole education. The variance teaches you your own business faster than any report: which customers pay later than they promise, which costs you underestimate, which weeks always run tighter than they look. After a month, your forecast gets genuinely accurate, not because you got better at predicting, but because you stopped guessing and started remembering.

Then it takes twenty minutes. Put it in the weekly meeting you’ve already built, right after the scoreboard.

Your First One Will Be Wrong

It will. Every first forecast is. Owners abandon the tool in week two because the numbers missed, and decide their business is too unpredictable to forecast.

It isn’t. The first version rests on assumptions you’ve never tested. The second is closer. By the fourth it’s usually close enough to act on, and you can say the date of your next tight week without opening anything.

Don’t forecast to the penny. Round to hundreds. Precision isn’t the point; direction and timing are.

The Test

• Can you name your lowest cash point in the next thirteen weeks: the number and the week it lands?
• Do you know your floor: the balance below which you cannot operate?
• Does one person own the file and update it every week, whether or not cash feels tight?
• Is it built from when money moves, or from when revenue is booked?

If the last answer is “when revenue is booked,” you have a P&L on a weekly grid. That’s a different tool, and it won’t warn you about anything.

A forecast doesn’t create a dollar. It buys you weeks, and weeks are the only thing that turns a cash problem into a cash decision.

If you want the row-by-row build sheet to work from, it lives at alinement.com/brief/tools/thirteen-week-cash-forecast.

Do you know your cash low point for this quarter, the week and the number, without opening your accounting system? Reply and tell me.

P.S. Next week: the AI tool. One page that answers where AI belongs in your business, who owns it, and what the guardrail is, so it stops being an experiment and becomes part of the work. The argument behind it is Issue 3, if you want the background first.