Building a Cash Flow Forecast for Uncertain Months

Why a fixed annual budget stops being useful

Most small practices start the year with a sensible budget: a target for fee income, a rough idea of costs, and a hope that the two meet somewhere pleasant in December. By the middle of February it is already out of date. A client delays a project, a big invoice sits unpaid for six weeks, and the tidy spreadsheet in the shared drive becomes a record of what you once believed rather than a tool you can steer with.

That is not a failure of discipline. It is the wrong instrument. A budget is a photograph taken in January. A rolling forecast is a moving picture, and for a practice of two to twenty people, the moving picture is what keeps you out of trouble. It also tells you when you can safely spend money, which is the part people forget.

What a rolling forecast actually looks like

You do not need software, an accountant on retainer, or a finance function. You need one spreadsheet with a row for each week or month, and columns for opening balance, money in, money out, and closing balance.

• Thirteen weeks, week by week. This is where late invoices bite, and where a weekly view shows you the exact week a payment gap appears.
• Twelve months, month by month. This is where investment decisions live — a new hire, a software switch, a lease renewal.
• The first week of every month, you replace the numbers with what actually happened and push the horizon forward by one month. The forecast never ends; it just rolls.

Keep it to a single file. The moment you have two versions, you will trust neither.

Start with the income you are confident about

Split expected income into three buckets. Committed means retainers already agreed, fixed-fee instalments already scheduled, and work in progress that is genuinely ready to invoice. Probable means quotes with a real client, a real date and a real budget. Speculative means everything else.

Only committed money goes in your base case. Probable work goes in the optimistic version, and speculative work goes nowhere near the forecast at all — it belongs in a business development plan, not in the numbers you rely on to pay salaries.

Then model the lag honestly. If you bill on 30-day terms and your clients typically take 45 days, work you deliver in March turns into cash in mid-May. Two-thirds of forecast errors in professional practices come from assuming clients pay when the invoice says they should, not when they actually do.

Build in the awkward costs

Regular costs are easy. It is the lumpy ones that catch people out, and they are entirely predictable if you list them once.

• VAT returns and payments, if you are registered
• PAYE and pension contributions, monthly or quarterly
• Quarterly rent, service charge and business rates
• Professional indemnity and other insurance renewals
• Corporation tax, and payments on account if you are self-employed
• Software subscriptions, annual licences and conference bookings
• Drawings or dividends, deliberately scheduled rather than taken as they come

Spread each one into the month it will actually leave the account. Suddenly the pattern becomes obvious: there are usually two or three months a year when everything lands at once, and those are the months to protect.

Plan for three versions, and set trigger points

Uncertainty is not a reason to avoid forecasting; it is the reason to forecast in ranges. Run three simple versions side by side:

• Base case — committed work only, standard payment behaviour.
• Slow case — the two largest invoices slip by two months and one client pauses their retainer.
• Good case — the probable work converts on time.

Now attach decisions to the lines rather than to the calendar. If cash falls below one month of operating costs, pause recruitment and defer non-essential spending. If it stays above three months, you have room to invest. Written down in advance, these rules stop being arguments and start being policy.

As a rough target for a small practice, holding eight to twelve weeks of operating costs in reserve covers most quiet periods without tying up capital you could be using.

Turn it into a ten-minute habit

A forecast that takes half a day will be abandoned by spring. Make it small and unglamorous.

• Download the bank statements and update actuals — ten minutes.
• Check the next six weeks specifically, not the whole year.
• Ask one question: what is the single decision this forecast is telling me to make?
• Write that decision down, and who owns it.

Then share the summary with your team. People plan their own lives around how the practice is doing; being straight with them about a quiet quarter builds far more loyalty than pretending it is not happening. A rolling forecast will not remove uncertainty from running a practice, but it turns a nasty surprise into a decision you made on purpose a month earlier. That is a much better place to be standing.