Most budgets are built by opening last year’s file and adjusting the numbers.

Revenue gets a growth percentage applied. Costs get inflated by a similar margin. A few line items get challenged by the CFO and revised down. The result is a plan that everyone knows is not quite right but cannot easily argue with, because every number has a precedent.

By March, the business has moved on. The budget has not.

The Problem with Incrementalism

Line-item budgeting is, at its core, a history exercise. You are asking: what did we spend last year, and how much more or less should we spend this year? The question is financially grounded but operationally disconnected. It assumes that the structure of last year’s costs is roughly the right structure for next year, which is often the assumption nobody has actually tested.

The deeper problem is what it does to the budget conversation. When assumptions are expressed as percentages applied to historical line items, it is very hard for a business unit head to challenge them meaningfully. They can argue that 8% is too aggressive and propose 5%, but neither number is connected to anything the business is actually doing. The conversation circles without landing.

Driver-based budgeting starts from a different question entirely. Not “what did we spend last year” but “what does the business need to do next year, and what does that require?” The difference sounds subtle. In practice it changes everything about how the budget is built and defended.

What a Driver Is and What It Is Not

A driver is an operational variable that has a direct, quantifiable relationship to a financial output.

Headcount is a driver. If you know how many people you are planning to hire and when, you can build salary costs, benefits, recruitment fees, equipment, and onboarding expenses from that single number. The financial line items follow. They are outputs, not inputs.

Revenue per customer is a driver. If you know your planned customer count, your average contract value, and your expected churn rate, you can build a revenue forecast from those three variables rather than from last year’s revenue number with a growth rate applied.

A percentage increase applied to a historical line item is not a driver. It is an assumption dressed up as a calculation. The difference matters because a driver can be challenged with business logic - “your headcount plan implies a productivity ratio we have never achieved” - whereas a percentage can only be challenged with negotiation.

The test for whether something is a driver is simple: can a non-finance person in the business tell you what moves this number, and why? If the answer is yes, you have a driver. If the answer requires you to explain the spreadsheet, you have a line item.

Building a Driver Tree

A driver tree maps the relationship between operational decisions and financial outcomes. Here is a simplified version for a 50-person business.

Revenue side

The top of the tree starts with the commercial assumptions:

  • Number of customers at the start of the period
  • New customer additions (from sales pipeline and conversion rate assumptions)
  • Churn rate (from customer success data and renewal history)
  • Average revenue per customer (from pricing and contract mix)

From those four variables, you can build monthly recurring revenue, annual contract value, and full-year recognised revenue. Every number in the revenue forecast traces back to an operational assumption that a salesperson, a product leader, or a customer success manager can speak to.

Cost side

The cost tree runs on similar logic:

  • Headcount by function, month of hire, and grade (drives salaries, benefits, office space, equipment)
  • Cost of goods as a percentage of Revenue (drives variable delivery costs)
  • Planned marketing spend by channel (drives campaign costs and the pipeline assumptions on the revenue side)
  • Fixed infrastructure costs (rent, software, utilities - these are genuinely fixed and can be taken directly)

The discipline is in separating costs that are genuinely fixed from costs that vary with business activity. Most businesses have fewer truly fixed costs than their line-item budget implies.

The Budget Conversation - This Makes Possible

The practical difference between a driver-based budget and a line-item budget shows up most clearly in the room when assumptions are being challenged.

In a line-item budget review, the CFO looks at the marketing spend line and asks why it is 15% higher than last year. The finance team explains the percentage. The marketing lead argues for a higher number. Nobody is talking about what the spend is expected to produce.

In a driver-based review, the same conversation goes differently. The marketing spend is connected to a pipeline assumption, which feeds a conversion rate, which produces a new customer number, which drives revenue. When the CFO challenges the marketing budget, the question becomes: if we reduce this spend by 20%, what does that do to the pipeline, and what does the pipeline do to revenue? The marketing lead now has to defend a business outcome, not a budget line.

This is what influence in the planning process actually looks like. The influence does not come from presenting a polished spreadsheet. It comes from being the person who built the logic that connects operational decisions to financial consequences, and who can trace any number in the plan back to an assumption that someone in the business owns.

Where Variance Analysis Connects

Driver-based budgeting and variance analysis are two ends of the same process.

When you build a budget on operational drivers, your monthly variance analysis becomes a diagnostic tool rather than an accounting exercise. A revenue miss does not just show up as a volume variance - it shows up as a deviation in a specific driver. Customer additions came in below the pipeline assumption. Churn ran higher than the renewal rate implied. Average contract value fell because of a mix shift toward smaller customers.

Each of those diagnostics points directly to an owner and a question: why did the driver move, and is the movement structural or situational? I covered the mechanics of that conversation in Variance Analysis: Making the Monthly Actuals Review Actually Useful. The two articles are designed to be read together because the budget and the actuals review are the same process viewed from opposite ends.

When Zero-Based Budgeting Makes Sense

Zero-based budgeting (ZBB) is the more radical version of the same instinct. Instead of building from last year’s base and justifying changes, every budget line must be justified from zero as if the cost did not previously exist.

ZBB is powerful in specific situations: a business that has grown rapidly and accumulated cost structures that no longer match its operating model, a company under significant margin pressure that needs to make genuinely difficult choices, or a new business unit being stood up for the first time. In those contexts, ZBB forces the prioritisation that incrementalism avoids.

The honest limitation is the cost of the process. A full ZBB exercise is resource-intensive, politically difficult, and disruptive to the finance team’s capacity for the rest of the planning cycle. For most businesses in most years, a driver-based approach achieves 80% of the analytical benefit at a fraction of the effort.

The question worth asking before committing to ZBB is whether the business has a cost structure problem or a cost visibility problem. Driver-based budgeting solves the visibility problem, it makes costs traceable to decisions. ZBB solves the structure problem, it forces a clean-sheet assessment of whether those decisions should be made at all. Most of the time, you need the first before you can do the second well.

A Note on Defending Assumptions

The most common failure in a driver-based budget is not the model - it is the assumptions that feed it.

A driver tree built on optimistic pipeline conversion rates, heroic productivity assumptions, or churn rates that have never been achieved in practice is not a driver-based budget. It is an incremental budget with extra steps. The rigour of the approach depends entirely on the rigour of the assumptions, and the assumptions are only as good as the conversation you had with the business before you built the model.

The best Finance BPs build the driver tree collaboratively, not in isolation. The sales leader owns the pipeline and conversion assumptions. The operations lead owns the delivery cost ratios. The people lead owns the hiring plan and attrition rate. The finance team’s job is to assemble those inputs into a coherent financial picture and identify where the assumptions are inconsistent with each other — where the hiring plan implies a headcount productivity ratio the business has never achieved, or where the revenue assumption requires a churn rate the customer success team says is not realistic.

That conversation, more than the model itself, is where the work actually lives.

Driver-based budgeting lays the foundation for the next layer of planning complexity: scenario analysis. Once your budget is built on explicit operational drivers, you can stress-test those drivers under different assumptions - a slower pipeline, a higher churn rate, a cost structure under pressure and understand the range of financial outcomes without rebuilding the model from scratch. I will cover that in Scenario Planning That Actually Gets Used.

I would love to hear how your planning process works right now, whether you are building on drivers already or still fighting with the percentage-increment model. Let’s connect.