Driver-Based Budgeting and Rolling Forecasts for UK SMEs
An annual budget is obsolete by March and defended until December. Driver-based rolling forecasts fix the accuracy problem — but only if you keep a fixed target to hold people to.
The traditional SME budget takes six weeks, consumes the leadership team's autumn, and produces a document nobody references after the first quarter's variances make it indefensible. The problem is not effort or accuracy — it is that a line-item budget cannot be updated coherently, so it is left to rot instead.
Line items versus drivers
| Approach | How the number is set | What happens when reality moves |
|---|---|---|
| Line item + % | Last year's salary cost plus 4% | No mechanism to update; the variance is explained, not modelled |
| Driver based | Heads × average cost × months, plus NIC and pension rates | Change one head and the plan, cash and headcount report all move together |
| Line item + % | Revenue up 12% because the board wants growth | Unfalsifiable, and costs never move with it |
| Driver based | Sales heads × ramped quota × win rate | The revenue plan is testable against pipeline every month |
Choosing drivers that survive contact with the business
- It must be something an operational manager already measures weekly, not a finance construct.
- It must move the number materially — if a 20% change shifts profit by less than 1%, it is detail, not a driver.
- It must be forecastable from something observable: pipeline, bookings, headcount plan, contracted backlog.
- There should be an owner outside finance who accepts the assumption as theirs.
Eight to twelve drivers covers the whole plan for most SMEs. Beyond that you are rebuilding the line-item budget with extra steps.
The monthly rolling cycle
- Close the month to a fixed timetable so the actuals are stable by working day five.
- Update actual driver values — heads, utilisation, units — before touching any financial line.
- Ask each driver owner one question: has anything changed in the next three months?
- Re-run the forecast, extend the horizon by one month, and compare against both the prior forecast and the fixed annual target.
- Report the change in full-year outturn since last month, and attribute it to specific drivers.
- Log any decision the re-forecast triggers, with the trigger stated.
The fifth step is the one that changes behaviour. A board that sees "full-year EBITDA outturn moved from £1.42m to £1.31m, of which £70k is utilisation running two points below plan" is having a management conversation. A board that sees a new set of numbers each month is not.
Keeping accountability when the forecast moves
The objection to rolling forecasts is legitimate: if the number keeps changing, nobody is held to anything. Resolve it structurally. The annual target is fixed at the start of the year and is what bonuses and commitments attach to. The rolling forecast is the current expectation and attaches to nothing except resource decisions. Report both, always, in the same table.
UK specifics to build into the driver model
- Employer's National Insurance and pension contributions as a percentage applied to the headcount driver, not a fixed cost line.
- The apprenticeship levy where the payroll threshold is crossed mid-year.
- VAT quarter payment dates modelled as cash events distinct from the P&L accrual.
- Corporation tax at nine months and one day after year end, with any R&D credit offset timed separately.
Doing this in MouCFO
MouCFO builds budgets from drivers with an AI first draft off your own actuals, holds the fixed annual target and the rolling forecast side by side, and produces the budget-versus- actual variance analysis with account-level commentary each month for the board pack.
Frequently asked questions
What is driver-based budgeting?
Building the budget from operational quantities — heads, utilisation, units, conversion rates — rather than from last year's figures plus a percentage. Financial outputs are then calculated from those drivers, so changing an operational assumption updates the whole plan consistently.
Does a rolling forecast replace the annual budget?
No. Keep a fixed annual target for accountability and incentives, and run a rolling 12-month forecast alongside it as the current best view. Confusing the two is the most common implementation mistake — people either stop being accountable or start gaming the forecast.
How often should an SME re-forecast?
Monthly, immediately after the close, extending the horizon by one month each time. A monthly cycle takes half a day once the drivers are built; a quarterly cycle is usually too slow to change any decision.