You invoice more in one month than ever before, and three weeks later the account is thinner than it has been all year. Nothing went wrong: two clients pay on thirty-day terms, an annual tool renewed on the 14th, and the money you earned and the money you hold sit on different calendars. A cash flow forecast shows you that gap while there is still time to act on it. The SBA makes the same point about managing money in a small business. Here is how to build one when you are the entire company.
Profit, cash flow forecast: two different questions
Profit asks whether the work paid for itself over a period. Cash asks whether you can meet what is due on Friday. Both can be true at once, and in any given week they rarely agree. The cause is single: an invoice is recorded when you raise it and felt when it is paid. Everything awkward about a solo business lives in that gap — payment terms, late payers, annual renewals that arrive whole, a thin month trailing a busy one by the length of your terms. It lands harder on one person because there is nothing else in the account.
A cash flow forecast does not fix any of that. It moves the discovery earlier, into a week where you can still change something.
The cash flow forecast is four rows and thirteen columns
Columns are weeks, not months: a renewal charges on the 14th, not “in June”, and a monthly grid hides the exact week that hurts. Rows are these four:
- Opening balance — what is actually in the account on Monday morning.
- Money in — every payment you expect to land that week.
- Money out — every payment that will leave that week.
- Closing balance — opening, plus in, minus out. It becomes next week’s opening.
That chain of closing-into-opening is the whole mechanism; everything else hangs beneath those four rows. Take invented figures, picked at random and meaning nothing about any real business: you open at 4,000, expect 1,800 in and 2,600 out, and close at 3,200. That 3,200 opens the next week, which takes in 900, pays out 3,100 and closes at 1,000 — a week you are reading now instead of living later.
Thirteen weeks is the working horizon: about a quarter, far enough to see a slow patch forming, near enough that you still know most of the amounts. A twelve-month version is a planning document: past the first quarter it is mostly guesswork, and guesswork you update weekly is anxiety with a grid.
Date every cash flow forecast line by when the money moves
One rule carries the accuracy of the whole sheet: a line goes in the week the money changes hands. Never the week the work happened, never the week the invoice was raised.
Money in: the payment date, not the invoice date
Start from the agreed terms, then correct for what the client actually does. If the last several invoices to one client were settled around day forty-five, enter day forty-five, not the thirty you agreed — an observation, not cynicism.
This depends on knowing what is coming. If the agreed work and the amounts live nowhere but your inbox, put them somewhere first — a simple CRM in Notion is enough, and the pipeline it holds feeds your money-in rows.
Money out: the debit date
Annual renewals catch people. In your head a yearly tool is a small monthly amount; in the account it is one charge on one day. Go through a year of statements once, list every annual and quarterly charge with its date, and drop each into the right week.
Mark each line with a confidence
Three levels are enough. Confirmed: signed, dated, or already invoiced. Expected: agreed in principle, not yet billed. Possible: a proposal out. Build the closing balances from confirmed and expected only, and keep the possible lines visible but out of the totals. The sheet then answers two questions at once — what happens if nothing new lands, and what changes if it does.
The weekly cash flow forecast update is the whole discipline
A rough forecast you touch every week is worth more than an accurate one you built in January. Twenty minutes, the same day each week, four moves:
- Reconcile. Set last week’s actual in and out beside what you predicted. Do not correct the history; the difference is what you want to see.
- Roll. Add a new thirteenth week at the end, so the horizon stays constant.
- Re-date. Move the lines that moved — a client said next Tuesday, so the line goes to next Tuesday.
- Write one sentence. What surprised you, and why. Everyone skips this, and it is the only step that teaches you anything.
Reconciling needs a clean record of what happened, which is a different job from forecasting. If you would rather not build one, The Cashflow Ledger is a Notion system for it: a categorised transactions ledger with a tax-deductible flag, an invoice board that turns a due date red the day it slips and computes days-to-payment per invoice, client totals that roll up on their own, and a recurring database with a next-due date that rolls itself forward and an annualised-cost column. Days-to-payment is what dates your money-in rows honestly; the recurring list holds the annual charges you forgot.
Runway in a cash flow forecast is arithmetic, not a prediction
Runway is cash divided by what leaves per period. Two thousand in the account against a thousand a week going out is two weeks, and it says nothing more than the division says. It assumes outgoings stay flat and nothing new arrives, which is the one thing you can be sure is wrong. That makes it comparative rather than useless: compute it the same way every week and read the direction, not the value.
Decide once whether the outflow figure is an average of recent weeks or the forecast for the coming ones, and whether expected income counts. Two runway numbers from two methods are not comparable, and the comparison was the point.
What a cash flow forecast can and cannot change
A cash flow forecast earns its keep in the weeks before a tight week, not during it. Seeing week nine dip while you are still in week two leaves you options, and they are all about timing:
- Invoice on delivery, not at month end. Batching to the last day of the month is a fortnight of delay you chose.
- Set terms and staging when the work is agreed. A deposit, a mid-point and a balance are easier to propose before a project than halfway through.
- Chase on a schedule, not by mood. A reminder the day after due, then a set interval, removes the hesitation that lets late invoices sit.
- Move what is movable. Renewals whose date you control, equipment, anything not committed to a day.
- Bring forward what is already earned. Work delivered and not yet billed is the fastest line to change.
Then the honest part. Everything past timing — borrowing, tax treatment, how much to hold back, what a contract entitles you to charge — is a conversation with an accountant or a lawyer, not a cell in a spreadsheet. Nothing here is financial or tax advice: rates, obligations and deadlines depend on your country and your setup, and none of it is settled in a grid of weeks.
The sheet will not make a quiet quarter busy either; it shows you the quiet quarter sooner, which is not the same thing. Nor will it tell you whether the business works — whether your pricing covers your costs, whether the model holds at three times the volume. Those are plan questions, and we went through them in what founders actually need from a business plan.
FAQ
How far ahead should a cash flow forecast go?
Thirteen weeks is the usual working answer: it covers a quarter while most of the amounts are still knowable. Keep a coarser monthly view beyond it if you like, but as a separate document, and do not update it weekly. Mixed together, the first month is a forecast and the eleventh is a wish, presented with identical confidence.
Do I need accounting software to forecast cash flow?
No — the two do different jobs. Accounting software records what happened and is built around your books; a forecast is a claim about what happens next and sits outside them. One person with a handful of clients can run thirteen weeks in a spreadsheet or a template, and the record feeds the forecast rather than replacing it.
How should tax appear in the forecast?
As an ordinary outgoing line: a date and an amount, placed in the week the money actually leaves. The same goes for a set-aside if you keep one. The forecast’s job is to hold that number where you can see it coming, not to work out what it should be — the figure, the deadline and what applies to you are questions for a qualified professional in your jurisdiction.
Go further
A thirteen-week forecast answers one question: can you meet what is due. The wider ones — what your margin actually is, how long the cash lasts at the current burn, what has to be true for next year — belong in a plan.
The Business Plan Studio is built around that: five plain numbers a month (revenue, fixed costs, variable costs, cash in the bank) and it computes gross margin, monthly burn and runway, with a plain-English explainer above the table. Around it sit nine guided plan sections with the questions an investor would ask, a worked example throughout, scored risks, and milestones whose progress computes itself. It runs on Notion’s free plan, as the Ledger does.