The thirteen-week cash forecast that fits on one page

A budget tells you what you hope will happen in a year. A cash forecast tells you whether you can pay people next Friday. For a company with fewer than a hundred people, the second tool matters more, and it is much easier to build.

The thirteen-week cash forecast is the oldest tool in finance, and it has not changed because it works. It covers one quarter, one row per week, and it fits on one page.

What goes on the page

Start with the bank balance at the beginning of the week. Below it, list the cash coming in: customer payments, grants, any money you expect from a funding round, and nothing else. Below that, list the cash going out: payroll, suppliers, rent, tax, loan payments, and the software invoices that are always forgotten.

The last line is the bank balance at the end of the week. It becomes the first line of the next week. Thirteen columns, one balance carried forward. That is the whole structure.

Timing is the whole point

Most forecasting errors are not about amounts. They are about dates. A customer who has a thirty-day contract rarely pays in thirty days. Use the real average from the last six months. If your customers pay in forty-four days, the forecast says forty-four.

Do the same with your own payments. Payroll lands on the same day each month. Taxes land on dates fixed by law. Suppliers are often paid whenever the finance person finds the time, and that is the one habit worth changing.

Three scenarios

One forecast is a guess. Three are a plan. Keep the base case as honest as you can. Then copy the page twice. In the slow case, move every customer payment back by two weeks. In the fast case, move it forward by one.

Look at the lowest balance in each scenario and the week it happens. If the slow case dips below one month of payroll, you do not have a forecasting problem. You have a financing problem, and you have thirteen weeks to solve it instead of three.

The weekly habit

Update the page every Monday. Replace last week's forecast with what really happened, shift the remaining weeks forward and add a new thirteenth week. This takes thirty minutes once the habit is set, and it replaces a good many difficult meetings.

The reward is quiet. When a customer pays late, nobody panics, because the forecast already shows what it changes. When an investor asks about cash, the answer is on one page.

If you build one finance tool this quarter, build this one. If you would like it built with you, that is a two-day job.