The Operational Model | Driver-Based Forecasting Guide

Here is a question that separates a junior analyst from a useful one: your commercial team wants to hire three more senior consultants in Q2. Each earns £70,000 per year, takes three months to become fully productive, and bills at £950 per day. Utilisation across the team is currently running at 72%. What happens to the P&L?

A model built on a single “revenue grows 8% this year” assumption cannot answer this. It does not know what drives revenue. It cannot distinguish between adding people and raising prices. It cannot tell you what happens to margin if utilisation slips by five percentage points during a hiring push. All it can do is compound last year’s numbers forward and hope the assumptions hold.

An operational model can answer the question precisely — in seconds, and with the ability to run a dozen variations before the meeting ends. This post builds one from scratch for ServiceCo, a fictional B2B professional services business, and in doing so covers the methodology that FP&A teams use across industries from consulting to SaaS to manufacturing.

What Is an Operational Model?

A standard P&L projection works top-down: you take last year’s revenue, apply a growth rate, then estimate costs as a percentage of revenue. It is quick to build and perfectly adequate for rough analysis. But it has a critical weakness — the assumptions are financial, not operational. They tell you nothing about the business decisions that actually determine whether those financial outcomes materialise.

An operational model inverts this logic. Every revenue and cost line is derived from explicit business drivers — the real-world inputs that commercial and operational teams actually control. The model is built bottom-up from those drivers, and the P&L is the output, not the starting point.

The hierarchy of drivers typically has three levels:

Volume drivers — the count of things: headcount, active customers, units sold, transactions processed
Rate drivers — the price or value attached to each unit of volume: day rate, average revenue per user, selling price per unit, cost per hire
Capacity drivers — the efficiency at which volume converts to output: utilisation rate, fill rate, hours available per FTE, machine uptime

In a well-structured operational model, every material line item traces back to at least one of these driver categories. When an assumption changes — say the commercial team revises its hiring plan, or a price increase is delayed by a quarter — the model propagates that change through to EBITDA automatically.

This transparency is one of the main reasons FP&A teams prefer driver-based models for internal planning. When the numbers change, you can explain exactly why. Non-finance stakeholders can challenge the assumptions because the assumptions are expressed in language they understand — headcount and utilisation rates, not revenue growth percentages.

Identifying Your Key Drivers

Before building anything, you need to identify the five to seven drivers that explain 80% of your revenue and cost. Every business has a different set, but there are recognisable patterns by business type:

Business type	Revenue driver	Key cost driver
Professional services	Headcount × Utilisation × Day rate	Salaries + overhead per head
SaaS / subscription	Active customers × ARPU	Headcount + hosting
Manufacturer	Units sold × Selling price	Materials + direct labour
Retailer	Footfall × Conversion rate × Average transaction value	COGS + store costs

For ServiceCo, the revenue drivers are straightforward:

Fee-earning headcount — how many consultants are in post (split by grade: Analyst, Consultant, Senior Consultant, Manager)
Utilisation rate — the percentage of available working days that are billed to clients (target: 75%)
Day rate — the billing rate charged to clients, which varies by grade

These three drivers, multiplied together, produce revenue. The model’s job is to make that multiplication visible and controllable.

A practical rule when identifying drivers: start with the commercial team’s operating plan, not the finance function’s spreadsheet. Commercial directors and heads of delivery typically think in exactly these terms — headcount, utilisation, rates. The operational model should speak their language.

Building the Revenue Model

The ServiceCo formula

The core revenue calculation for a professional services business is:

Revenue = Headcount × Available Days × Utilisation Rate × Day Rate

Where Available Days is the number of working days in the period minus holidays and non-billable training time. For a standard UK year this is approximately 220 days per FTE.

For ServiceCo, the grade structure and rate card look like this:

Grade	Headcount (Year 1)	Day Rate (£)	Utilisation Target
Analyst	12	600	80%
Consultant	8	800	75%
Senior Consultant	5	950	72%
Manager	3	1,200	65%
Total	28	—	—

Applying the formula to the Senior Consultant grade: 5 × 220 × 72% × £950 = £748,440 of annual revenue from that grade alone.

Summing across all grades produces total fee income. The full Year 1 revenue calculation for ServiceCo:

Grade	FTE	Days	Utilisation	Day Rate	Revenue (£k)
Analyst	12	220	80%	£600	1,267
Consultant	8	220	75%	£800	1,056
Senior Consultant	5	220	72%	£950	748
Manager	3	220	65%	£1,200	514
Total	28				3,586

Building it in Excel

The Excel build follows a logical sequence:

Headcount Plan tab — opening headcount by grade, plus monthly hires and leavers. A rolling sum gives closing headcount each month; the average of opening and closing gives the FTE for the period. This is important: a consultant hired in October contributes only three months of revenue in Year 1, not twelve.
Assumptions tab — the rate card (day rate by grade, updated annually for inflation), utilisation targets by grade, and available working days per year. All inputs live here. Nothing is hardcoded in formulas elsewhere.
Revenue Model tab — applies the formula: =Average FTE × Available Days × Utilisation × Day Rate. Built monthly, then aggregated to quarterly and annual totals for the P&L summary.

The result is a revenue model where changing a single utilisation assumption — say, dropping Senior Consultant utilisation from 72% to 65% during a major pitch period — immediately flows through to the revenue line and then to EBITDA.

Building the Cost Model

Costs in a services business divide cleanly into three categories: direct (vary with headcount), semi-variable (step up at capacity thresholds), and fixed (largely independent of volume).

Direct costs

These link directly to the headcount plan:

Consultant salaries — the single largest cost. Pull the grade-level headcount from the Headcount Plan tab and multiply by salary by grade. Include employer’s National Insurance and pension contributions as a percentage uplift (typically 14–16% for UK employers).
Subcontractor spend — when utilisation runs above target or a specialist skill is needed, ServiceCo brings in contractors. Model this as a percentage of revenue above a utilisation threshold.

Semi-variable costs

These do not scale linearly with headcount but jump at certain thresholds:

Office space — ServiceCo’s current lease covers 35 desks. Once headcount exceeds 35, a desk-sharing ratio kicks in; once it exceeds 45, a new lease is required. Model this as a step function: =IF(Headcount<=35, Base_Rent, IF(Headcount<=45, Base_Rent × 1.2, Base_Rent × 1.6)).
Technology licences — per-seat pricing for project management tools and modelling software. Cost scales exactly with headcount.
Recruitment costs — a fee (typically 15–20% of first-year salary) triggered only in periods where hires are planned.

Fixed overhead

Leadership, finance, HR, and central IT costs are broadly fixed in the short term and do not respond to changes in fee-earner headcount. Budget them as annual figures and spread evenly across the year.

Cost driver mapping for ServiceCo

Cost line	Driver	Behaviour
Consultant salaries	Grade headcount × salary	Variable
Employer NI & pension	Salary bill × 15%	Variable
Subcontractor	Revenue above utilisation cap	Variable
Office rent	Headcount threshold	Step-fixed
Technology licences	Total headcount	Variable
Recruitment fees	Planned hires × avg. salary × 18%	Variable
Senior leadership	Fixed budget	Fixed
Finance & HR	Fixed budget	Fixed
Training & CPD	Fee-earner headcount × £800/head	Variable

The Full Excel Tab Structure

A clean operational model for ServiceCo has six tabs:

Assumptions — all rate cards, salary scales, utilisation targets, working day calendars, and overhead budgets. This is the single source of truth. If a number appears more than once in the model, it should appear here once and be referenced everywhere else.

Headcount Plan — the monthly hiring and attrition schedule. Build it as: Opening + Hires − Leavers = Closing. Use helper rows to compute the monthly average FTE, which feeds the Revenue Model and the Cost Model.

Revenue Model — grade-level revenue calculation (FTE × Days × Utilisation × Rate), aggregated to monthly and annual totals.

Cost Model — each cost line with its driver clearly labelled. Direct cost lines reference the Headcount Plan; step-fixed lines use IF logic from Assumptions; fixed overheads are hardcoded annual figures spread monthly.

P&L Summary — Revenue − Direct Costs = Gross Profit → − Overhead = EBITDA → − D&A = EBIT → tax and interest bridge to net profit. Include a FCF bridge at the bottom: EBITDA − CapEx − Working Capital movement.

Scenario Switch — a simple dropdown or toggle that overwrites the central assumptions. Bear case: utilisation drops to 65%, one round of redundancies. Bull case: utilisation holds at 80%, two additional senior hires in H2. The P&L Summary updates in real time.

Key Excel techniques that make this work cleanly:

Named ranges for rate card items: Day_Rate_SC rather than $C$14. The model becomes self-documenting.
OFFSET with a month counter for dynamic monthly references, avoiding the need to manually update cell addresses as you copy formulas across columns.
Conditional formatting on the utilisation row — red if below 65%, amber between 65–70%, green above 70%. Provides an instant visual check on model health.

How FP&A Teams Use This Model

The operational model is not a once-a-year exercise. In a well-run finance team, it is refreshed every month as part of the reforecast cycle.

The mechanics are straightforward: at the start of each month, actuals replace the plan assumptions for all periods that have closed. Headcount is updated with confirmed hires and leavers. If a major client engagement ended earlier than expected, utilisation is revised downward for the affected grades. The remaining months of the year are replanned from the updated position.

What makes this valuable is not precision — the future is always uncertain — but speed and transparency. When the business misses its revenue target in March, a driver-based model tells you whether it was a volume problem (lower utilisation than planned), a rate problem (day rates being discounted to win work), or a mix problem (a different grade composition than assumed). A top-down model cannot make that distinction; it only tells you the revenue was £X lower than budget.

This diagnostic capability is what commercial finance teams are hired to provide. The operational model is the tool that makes it possible.

Key Takeaways

Driver-based models are built bottom-up from operational reality — headcount, utilisation, pricing — rather than top-down from financial growth rates
Identify the five to seven drivers that explain 80% of revenue and cost before building a single formula
For a professional services business, the headcount plan is the central driver; every other number derives from it
Structure the Excel model so every assumption lives in one tab and every output formula references it — this makes reforecasting fast and auditable
The operational model’s real power is diagnostic: when actuals diverge from plan, it tells you exactly which driver moved and by how much

Practice

Build the ServiceCo operational model for a two-year period using the grade structure and rate card above. In Year 2, add two Consultant hires in April and one Senior Consultant hire in September. Then stress-test the model by reducing utilisation from 75% to 65% across all grades for the full Year 2. Trace the impact through to EBITDA and explain in plain language what drove the change. This scenario — a hiring push that temporarily depresses utilisation before new staff are fully productive — is one of the most common questions asked in FP&A interviews.

The Operational Model: Driver-Based Forecasting for Finance Teams