Driver-Based Planning and Rolling Forecasts
The planning discipline that builds numbers from the operational drivers that actually move them - volumes, rates and capacity - and refreshes them on a rolling horizon, instead of anchoring next year to last year plus a percent.
Driver-based planning builds the financial plan from the operational quantities that cause cost and revenue - units sold, transactions processed, orders shipped, minutes of work per task, and the capacity available to do it - rather than from prior-year figures adjusted by a flat percentage. Each line in the plan becomes an equation: revenue is price times volume, resource cost is activity volume times a time-per-unit times a cost rate, and headcount is workload divided by capacity. A rolling forecast then re-runs those equations every month or quarter over a constant horizon (typically the next twelve to eighteen months), so the plan always looks the same distance ahead and reacts as drivers change. The payoff is a plan you can explain, flex and re-forecast quickly, because every number traces back to an assumption you can defend. Its close cousin is the time-driven activity-based costing time equation, which is precisely the driver model that connects planned volumes to the capacity and cost they consume.
Plan the drivers, let the money follow
Traditional budgeting starts with last year's numbers and negotiates a change: sales up 5%, costs held flat, a headcount added here. The figures look precise but the logic is hidden, so when reality diverges nobody can say which assumption broke. Driver-based planning inverts the order. It asks what physically drives each number - how many customers, how many orders, how many minutes per order, how many people per thousand orders - and expresses the financial line as a function of those drivers.
Once the plan is written as equations rather than fixed totals, two things change. Re-planning becomes arithmetic: change the volume assumption and every dependent line recomputes, so scenario analysis is a matter of changing one input rather than rebuilding a spreadsheet. And accountability sharpens, because a variance now points at a specific driver - volume was higher, the rate per transaction slipped, capacity was under-used - instead of a vague "we missed budget". The plan stops being a target to defend and becomes a model to learn from.
Rolling forecast versus the annual budget
The annual budget fixes a target once a year and measures against it for twelve months. By month nine it is comparing today's decisions to a set of assumptions made fifteen months earlier - useful for control, poor for steering.
The rolling forecast keeps a constant horizon. Each period you drop the month just closed and add a new month at the far end, so you are always looking, say, eighteen months ahead. Because it is rebuilt from current driver values, it absorbs new information - a demand shift, a price move, a capacity constraint - instead of ignoring it until the next budget cycle.
The driver model is what makes rolling cheap. If a forecast is a hand-built number, re-forecasting monthly is unbearable work. If it is a driver equation, re-forecasting is updating the volume and rate inputs and letting the model recompute. That is why driver-based planning and rolling forecasts travel together: the first makes the second affordable, and the second makes the first worth doing.
Capacity closes the loop. A driver plan does not just forecast cost; it forecasts the resources needed to serve the planned volume, and flags where practical capacity is exceeded or left idle - the same capacity logic that sits at the heart of time-driven costing.
Planning a service team from its drivers
A back-office team processes two work types (illustrative figures, not client data). Standard orders take 12 minutes each; complex orders take 30 minutes. Next year's plan expects 180,000 standard orders and 40,000 complex orders. The fully-loaded cost of the team is €0.60 per available minute, and each full-time employee supplies 100,000 practical minutes a year.
Required minutes are (180,000 × 12) + (40,000 × 30) = 2,160,000 + 1,200,000 = 3,360,000 minutes. Planned resource cost is 3,360,000 × €0.60 = €2,016,000. Headcount needed is 3,360,000 / 100,000 = 33.6 FTE, so the plan calls for 34 people carrying a small buffer. Now flex a driver: if complex orders grow to 55,000, required minutes rise to 2,160,000 + 1,650,000 = 3,810,000, cost rises to €2,286,000 and headcount to 38.1 FTE - a €270,000 and roughly 4-FTE swing that traces to one assumption. The annual budget would have hidden that sensitivity inside a single cost line; the driver model surfaces it, and the rolling forecast would catch the shift the moment the volume trend appeared.
Where driver-based planning holds, and where it bends
| What driver-based planning relies on | Why it can break |
|---|---|
| The right drivers are identified for each line | A weakly correlated driver produces a confident-looking plan built on the wrong cause; the driver set needs periodic testing |
| Rates and times per unit are reasonably stable and measured | Learning curves, mix shifts and process change move the rates; stale coefficients quietly bias the plan |
| Volume forecasts are credible | Driver logic amplifies a bad volume assumption - garbage in, precisely computed garbage out |
| The organisation can re-forecast on a cadence | Rolling forecasts demand discipline and light tooling; without them the cadence lapses back to an annual scramble |
| Capacity is expressed at a practical level | Planning at theoretical capacity understates resource needs and hides the cost of idle time |
None of this makes the method fragile; it makes it a model that has to be maintained. Its equations are only as good as the drivers, rates and volumes behind them, and its rolling cadence only helps if the organisation keeps re-running it. That is the bridge to the rest of this encyclopedia: the time-per-unit and capacity coefficients come straight from time-driven activity-based costing, the volume and mix assumptions are sharpened by customer- and product-profitability analysis and cost-to-serve, and the profit consequences of a shifting mix are exactly what the whale curve makes visible.
When driver-based planning earns its keep
Strengths. It makes the plan explainable and flexible: every number traces to an operational assumption, scenarios are a change of input rather than a rebuild, and variances point at a specific driver instead of a generic miss. Paired with a rolling forecast it keeps the plan current, links financial numbers to the capacity that delivers them, and turns budgeting from an annual ritual into a continuous steering tool.
Limits. It costs effort to set up and maintain - drivers must be chosen, rates measured, and the model refreshed on cadence. Over-engineered, it drowns in drivers nobody trusts; under-maintained, its coefficients drift out of date. Treat it as a living model, keep the driver set lean and the rates current, and reserve the detail for the lines where volume and mix genuinely move the result.
Common questions about driver-based planning and rolling forecasts
- What is driver-based planning?
- It is planning that builds each financial line from the operational drivers that cause it - volumes, rates and capacity - so revenue becomes price times volume and resource cost becomes activity volume times a time-per-unit times a cost rate. Instead of anchoring next year to last year plus a percent, the plan is a set of equations you can flex by changing an assumption.
- How is a rolling forecast different from an annual budget?
- An annual budget fixes a target once and measures against it all year, so late in the year it compares decisions to assumptions made many months earlier. A rolling forecast keeps a constant horizon - dropping the closed period and adding a new one each cycle - and rebuilds from current driver values, so it always looks the same distance ahead and absorbs new information as it arrives.
- How does driver-based planning connect to TDABC?
- Time-driven activity-based costing expresses resource consumption as a time equation: minutes per unit of work times a cost per minute. That equation is exactly the driver model a plan needs. TDABC supplies the time-per-unit and practical-capacity coefficients, and driver-based planning runs them forward against planned volumes to forecast cost and headcount.
- Why plan at practical capacity rather than theoretical capacity?
- Practical capacity reflects the minutes a resource can realistically supply after breaks, maintenance and normal downtime - usually around 80 to 85% of theoretical. Planning at theoretical capacity understates the resources a given volume needs and buries the cost of idle time, so a plan built on practical capacity is both more honest and more useful for sizing headcount.
- Do rolling forecasts replace the annual budget entirely?
- Not always. Many organisations keep an annual target for governance and incentives while using the rolling forecast to steer. The point is to stop treating a twelve-month-old budget as the live view of the future; the rolling forecast becomes the operational plan, and the budget, if kept, becomes a reference point rather than the daily guide.
References
Kaplan, R. S. & Anderson, S. R. Time-Driven Activity-Based Costing (time equations and practical capacity as the driver model). · Horngren, C. T., Datar, S. M. & Rajan, M. V. Cost Accounting: A Managerial Emphasis (budgeting, flexible budgets and cost drivers). · Kaplan, R. S. & Norton, D. P. The Strategy-Focused Organization (linking operational drivers to financial plans). · CIMA, Official Terminology (definitions of budget, forecast and cost driver). · IMA (Institute of Management Accountants), Statements on Management Accounting (planning, budgeting and driver-based approaches).