.png)
Key Takeaway
Driver-based forecasting replaces line-by-line guesswork with a small set of business drivers, such as volume, price, headcount and utilisation, that explain most of your results. When the drivers change, the forecast updates, and the conversation moves from "what's the number?" to "what's changing it?"
Most budgets are built the same way every year. Take last year's actuals, add a percentage, argue about it, and lock it in. The result is a forecast that looks precise and goes stale within a quarter.
Driver-based forecasting takes a different route. Instead of forecasting hundreds of lines, you identify the handful of operational factors that actually move your numbers and build the plan from them.
If you can explain why a number moved, you can forecast where it's heading. If you can't, you're guessing with extra steps.
A driver is an operational measure that has a direct, explainable link to a financial result. Revenue doesn't just appear. It comes from units sold, price, customer count or billable hours. Costs follow something too: headcount, production volume, shipments, contract terms.
Examples of Drivers by Line Item
| Financial Line | Typical Drivers | Owned By |
|---|---|---|
| Revenue | Units sold, average price, active customers, churn rate | Sales and Commercial |
| Cost of Goods | Production volume, material cost per unit, yield | Operations |
| Personnel | Headcount, average salary, hiring dates, attrition | HR and Department Heads |
| Services Delivery | Billable hours, utilisation, rate card | Delivery and Project Leads |
| Working Capital | Days sales outstanding, inventory days, payment terms | Finance and Supply Chain |
The FP&A Trends Group's 10th annual survey (475 finance professionals, sponsored by OneStream) shows the strain many teams are under. It's a vendor-sponsored survey and the results are self-reported, so read them as directional, but the picture is consistent with what we hear from finance teams.
Source: FP&A Trends Group, 10th Annual FP&A Trends Survey (2026), 475 respondents, as published by OneStream, the survey sponsor. Self-reported.
When nearly half the team's time goes to collecting and cleaning data, there's little left for the analysis that makes a forecast useful. And a plan built on last year's percentages can't answer "what if volume drops 10%?" without a week of rework.
Start with the Few Drivers That Matter
Ask which five to ten measures explain most of your revenue and cost. If a driver doesn't change a decision, leave it out. More drivers does not mean a better model.
Link Each Driver to a Financial Outcome
Define the relationship clearly. Units multiplied by price gives revenue. Headcount multiplied by loaded cost gives personnel expense. Keep the logic simple enough that a business leader can follow it.
Give Each Driver an Owner
The people closest to the operation should own the assumptions. Sales owns volume, operations owns yield, HR owns hiring dates. Finance owns the model and the challenge, not every input.
Refresh Drivers, Not the Whole Plan
When conditions change, update the driver and let the forecast recalculate. This is what makes rolling forecasts and what-if scenarios practical instead of painful.
Review Variances Through the Drivers
When actuals miss the forecast, ask whether volume, price or cost per unit moved. This turns a variance report into a conversation about what to do next.
A good driver-based model doesn't predict the future. It shows you quickly what changes when the future turns out different from the plan.
| Mistake | Why It Hurts | Better Approach |
|---|---|---|
| Too many drivers | The model becomes slow, hard to maintain and hard to trust | Limit it to the drivers that change decisions |
| Finance owns every input | The business doesn't feel accountable for the numbers | Assign driver ownership to operational leaders |
| Drivers that can't be measured | You can't track actuals against assumptions | Choose drivers that already exist in source systems |
| Rebuilding from scratch | Long projects lose momentum and stakeholder support | Start with one area, prove it, then expand |
Pick the area of the business where forecasts miss most often and the drivers are easiest to see. For many companies that's revenue, headcount or a single product line. Build a simple driver model, run it alongside your current forecast for a cycle or two, and compare. Teams are usually convinced less by the theory than by the first time a scenario takes minutes instead of days.
Technology matters here too. Driver-based models are hard to maintain across linked spreadsheets, which is why most teams eventually move them onto a planning platform where drivers, hierarchies and scenarios sit in one model.
Keansa helps finance teams identify the right drivers, design models the business can understand, and build them on enterprise planning platforms so forecasts update as conditions change. We start small, prove the value in one area, and expand from there.
Want to see what a driver-based model could look like for your business?
Talk to a Keansa Consultant.png){kind=small}
.png){kind=small}
.png){kind=small}