A fleet average gives you a temperature reading. It tells you whether the number is roughly where you expected it to be. What it cannot tell you is which vehicles are pulling the average up, why, and what you should do about it. For a 200-vehicle van fleet, that distinction is the difference between cost control and cost monitoring.
This article makes the case for VRM-level cost reporting as the practical standard for mid-sized commercial van operations — not as a technology project, but as a change in how cost data is structured and reviewed.
Why VRM is a more stable unit than driver
The instinct in fleet management is often to analyse cost at the driver level — linking fuel spend, incident rate, and maintenance events to named individuals. Driver-level reporting has genuine value for duty-of-care, insurance, and performance management. For cost reporting, however, vehicle registration mark is a significantly more stable analytical unit.
The reason is simple: vehicles persist; drivers turn over. In a 200-van fleet, an annual driver turnover rate of 20–30% is not unusual. A driver who leaves the business takes their name out of your driver-level cost report. The vehicle they were driving stays in your fleet and carries the cost consequences of how it was operated — worn brakes, a degraded DPF, tyres that were replaced early — forward into the next reporting period, now attributed to whoever drives it next.
Driver-level reporting treats the cost as having left with the driver. VRM-level reporting correctly keeps the cost with the vehicle. Over a 12-month view, the vehicle's cumulative cost history is the factual record; the driver attribution is a derived field useful for HR purposes but not the primary cost signal.
This is particularly true for pool vehicles and multi-shift operations, where a single VRM may be driven by four or five different individuals over a week. Driver-level cost attribution in these scenarios is essentially noise. VRM-level cost is always clean.
The four cost categories that need to be per-VRM to be useful
Not every cost category benefits equally from per-VRM tracking. The ones that do are the ones with meaningful vehicle-to-vehicle variance:
- Fuel: Variance between vehicles on comparable routes is the most immediate signal of driver behaviour issues, vehicle condition problems, or route inefficiency. Fleet-average fuel cost hides this variance almost entirely.
- Repair and maintenance: R&M per VRM is the clearest indicator of vehicle health trajectory. A van whose R&M cost is trending upward over three consecutive quarters is telling you something; a van at flat cost is telling you something different. That signal is invisible in a fleet aggregate.
- Depreciation: As covered separately in the residual value article, per-VRM depreciation based on actual market value movement is materially different from a straight-line book calculation spread across the fleet. Disposal decisions made against fleet-average depreciation are often made with incorrect assumptions about specific vehicles.
- Tyre spend: Tyre replacement frequency per VRM is a leading indicator of both driving behaviour and vehicle alignment or suspension issues. A vehicle replacing tyres at twice the rate of comparable vehicles on comparable routes has a problem — with the driver, with the vehicle, or with the route surface. Fleet-average tyre spend gives no view into this.
Costs that do not vary meaningfully between vehicles — fixed insurance premiums, operator licence fees, VED — can reasonably be apportioned by mileage fraction rather than tracked per VRM individually. The goal is to concentrate per-VRM detail where variance is actionable, not to create per-VRM tracking for its own sake.
How the reporting cadence works in practice
Monthly is the right primary cadence for VRM-level cost reporting in a mid-sized van fleet. Weekly is too granular for most cost categories — a single large R&M event in one week distorts the picture without enough subsequent data to contextualise it. Quarterly is too infrequent to catch developing problems before they compound.
A useful monthly VRM cost report for a fleet manager running 200 vehicles covers:
- Per-VRM cost this month, sorted descending (high to low) — the top 10% of vehicles by cost is the immediate management focus
- Month-on-month variance per VRM — a vehicle that was average last month and is now in the top decile has changed; a vehicle consistently in the top decile is a structural issue
- Rolling 3-month per-VRM cost — smooths one-off events and shows trajectory
- Cost-per-mile per VRM for fuel and R&M — normalises for mileage differences between vehicles
For the Finance Director, the same data generates a different cut. The FD wants fleet-total cost by category, variance against budget, and the distribution of cost across the fleet — specifically, what proportion of total cost is concentrated in what proportion of vehicles. In well-managed fleets, the rule of thumb is that the top 15–20% of vehicles by cost account for 35–45% of total variable cost. If your top 20% is accounting for 55% or more, you have a concentration problem worth investigating.
The FD cut versus the fleet manager cut
These are not two separate reports — they are two views of the same underlying per-VRM data. The fleet manager view sorts by vehicle; the FD view aggregates by category and shows distribution. Both are derived from a single per-VRM cost dataset. This matters because the most common reason fleet operators lack VRM-level reporting is not a data availability problem — it is a data structure problem. The underlying data exists in the fuel card export, the workshop invoicing system, and the telematics odometer feed. What is missing is the assembly step that joins it per registration mark.
Common objections — and where they hold up
Two objections come up regularly when fleet managers consider moving from fleet-average to VRM-level reporting.
The first is data quality: "our workshop records don't reliably use VRM as the primary key." This is a valid concern and a genuine implementation challenge. Workshop jobs are often filed against a job number or a customer account, with VRM as a secondary field that is sometimes missing or inconsistently formatted. Cleaning this retroactively is labour-intensive. The right response is to set VRM as a mandatory field in your workshop management system going forward, accept that the first two to three months of per-VRM reporting will have gaps, and use the gap analysis to pressure the data quality issue rather than waiting for perfect data before starting.
The second objection is administrative overhead: "tracking 200 vehicles individually will take more time than the insight is worth." This conflates building the report with reading the report. Once the data feeds are automated — fuel card export, telematics mileage, and workshop invoices all flowing into a joined per-VRM view — the monthly report generation is a system task, not a manual one. The fleet manager's time is spent reading the output and acting on the top-decile vehicles, not assembling the data. The setup cost is real; the ongoing overhead is low.
We're not saying fleet-average reporting is useless — a single headline figure is still valuable for board-level communication and budget variance tracking. The argument is that average-only reporting, as the primary cost management tool for a 200-vehicle fleet, leaves actionable information permanently buried. Per-VRM reporting does not replace the average; it makes the average meaningful by showing you what is inside it.
The practical starting point is identifying the 20 highest-cost vehicles in your fleet right now — not by driver name, not by depot, but by VRM — and asking whether you know why they cost what they cost. If the answer requires more than 10 minutes of data retrieval, your current reporting structure is the problem.