How many months of cash do we have left, and what would stretch it?

Revenue. A subscription software company at the end of September 2026: 2,795 customers on Starter at €29 a month, 1,230 on Growth at €99 and 467 on Scale at €299, €342,458 of recurring revenue a month. Its revenue comes from the model behind the revenue forecast demo, run 1,000 times from that month: each run draws its own churn level, then sign-ups, trials that pay, moves between plans and customers who leave or come back, month by month for three years.

Spending. 40 people at €8,000 a month each, fully loaded (salary, the employer’s taxes, benefits and equipment), and €110,000 a month for everything else: marketing, offices, software and advisers. New people join evenly over the next year, and the team then holds. A spending cut counts from October 2026 and stays.

Each month. Cash at the end of a month is cash at the start, plus 80% of the month’s revenue, less the month’s spending, from €4,200,000 in the bank at the end of September 2026. The 80% is gross margin, the cohort analysis demo’s: hosting, support and payment fees take the rest. A run runs out of cash in the first month its balance ends below zero, and the answer is the month by which half the runs have.

Six more months. The spending cut is the smallest, in steps of €1,000 a month from October 2026, that moves that month six months later. It is read off the same runs, so it moves with every control.

Lower churn. Each run is simulated twice from the same random draws, as now and with Starter churn one point lower, so the gap in cash within a run is the lower churn’s effect alone. A customer kept adds €29 a month at a time: at the settings the page opens with, lower churn adds €126,900 by May 2028, less than the company spends in that one month (€526,000).

What this shows, and what it does not

What it does not. The forecast counts recurring revenue as cash in the month it is billed: no annual plans paid up front, no late payers, no VAT or tax, no interest on the cash in the bank. Spending holds flat apart from the hires and the cut, with no pay rises and no one-off costs, and sign-ups stay at the model’s 600 a month whatever marketing spends. The range covers chance and the churn rate, nothing else: a large customer lost or a price change falls outside it.

On your own data. The revenue model is fitted to your billing history, and the spending comes from your ledger and your hiring plan. Each month the forecast is set against what happened and the inputs are updated; the method is on revenue forecasting. Anything you act on, such as a hiring plan or a cost cut, carries the analytics director’s sign-off.