Most van fleet reports give you one number: an average pence-per-mile figure across all vehicles. That figure goes into your monthly board pack. Your FD sees it. You nod and move on. The problem is that fleet-average pence-per-mile tells you very little about where your money is actually going — because the outliers on either end cancel each other out before the number ever reaches the page.
This guide works through the real TCO formula for a UK van fleet, explains why averaging destroys the signal, and describes a practical way to cut the data so it becomes a cost-control tool rather than a reporting ceremony.
The actual TCO formula — and what most systems miss
Per-mile total cost of ownership for a commercial van is:
TCO per mile = (Fuel + Repair & Maintenance + Insurance + Depreciation + Operator overhead) ÷ Total mileage in period
Each of those five elements has a different data source, a different update frequency, and a different owner inside your business. Fuel comes from your fuel card provider. R&M comes from your workshop management system or invoice archive. Insurance is typically an annual premium spread across the fleet. Depreciation requires a residual value estimate — which most operators either ignore or approximate with a straight-line formula that bears no relation to actual used-van market prices. Operator overhead covers operator licence fees, Vehicle Excise Duty, and administrative time that rarely gets apportioned per vehicle.
The gap between this formula and what most fleet management systems actually produce is almost always in depreciation and overhead attribution. Fuel and R&M get tracked tolerably well. Depreciation in particular tends to be either absent or averaged across a vehicle class rather than modelled per VRM.
Why fleet-average masks the problem vehicles
Consider a 180-van operation running a mix of short-wheelbase panel vans on urban routes and long-wheelbase Transits doing motorway-heavy distribution. The urban short runs will accumulate fuel cost at a higher pence-per-mile rate due to stop-start driving cycles. The motorway vehicles will have lower fuel cost per mile but higher depreciation in absolute terms due to mileage accumulation speed. If you average all 180 together, both patterns disappear.
More critically: within any route cohort, you will typically have a spread of roughly three driver behaviour classes. For a fleet running at a mean of, say, 24p/mile all-in on fuel alone, the low-cost cohort might run at 20–21p, the average cohort at 23–25p, and the high-cost tail at 28–32p or beyond. The high-cost tail vehicles are almost never identifiable from a fleet-average report. They are statistically submerged.
The three driver cost profiles
This is not a moral judgement about individual drivers — it is a structural observation about how cost distributes across a working fleet. We're not saying high-cost drivers are necessarily careless; route characteristics, vehicle age, and load profiles all contribute. But the pattern is consistent enough to be useful:
- Economy operators — typically 15–20% below fleet mean on fuel; R&M intervals run slightly longer than average; tyre wear within expected range for mileage.
- Average operators — cluster around the fleet mean; not a control problem but also not the source of savings.
- High-cost operators — often 25–40% above mean on fuel; R&M events cluster closer together; tyre replacements arrive ahead of mileage schedule. This cohort typically represents 10–15% of vehicles but can account for 20–25% of total variable cost.
The figure ranges above are illustrative and will vary by fleet type, vehicle mix, and route geography. The point is not the specific percentages — it is that the distribution exists and is stable enough to act on, once you have per-VRM data rather than averages.
How overhead attribution distorts the picture further
Vehicle Excise Duty, operator licence renewal fees, and DVSA compliance costs are real fleet costs that almost no per-mile TCO calculation properly assigns to individual vehicles. Most operators either ignore them entirely or divide the annual total by the fleet count to produce a per-vehicle flat rate.
That flat-rate approach is reasonable as a starting point, but it breaks down when your fleet has significant age and mileage variance. A 2021-plate Ford Transit Custom running 40,000 miles per year is not carrying the same cost profile as a 2018-plate Vauxhall Vivaro running 15,000 miles. Spreading a flat overhead rate across both produces a number that is wrong for both.
The practical fix is to apportion overhead by mileage fraction: if vehicle BV21 KXG covers 8% of total fleet mileage in a given period, it should carry 8% of the non-vehicle-specific overhead costs for that period. It is not a perfect method — some overhead costs are genuinely per-vehicle rather than per-mile — but it is substantially more accurate than a headcount split.
Getting the depreciation component right
Depreciation is the cost element most often either absent or wrong in fleet TCO calculations. It matters because for a van with a typical purchase price of £22,000–£30,000 and a four- to five-year life cycle, depreciation commonly represents 20–30p per mile on a fully loaded basis — comparable to or exceeding fuel cost on many routes.
Most operators use straight-line depreciation to book value: purchase price minus a fixed percentage per year. That method works for accounting purposes but does not reflect actual residual value movement, which for commercial vans follows a steeper initial decline in years one to three and a plateau through years five to seven. The difference between book value and actual market value at point of disposal is often a meaningful surprise.
Referencing CAP HPI or Glass's Guide valuations periodically — even quarterly rather than at disposal — gives a more accurate running depreciation figure to fold into your TCO calculation. It also allows you to spot vehicles where actual market value has diverged significantly from internal book value, which is useful information when making hold-versus-dispose decisions.
Building a TCO report that is actually usable
The test of a per-mile TCO report is whether you can read it and immediately identify which vehicles to act on. A fleet-average report fails that test by design. A vehicle-level report passes it — but only if the data inputs are consistent and the cost categories are genuinely comparable across VRMs.
Practical minimum requirements for a useful per-VRM TCO report:
- Fuel cost sourced from fuel card data reconciled against telematics odometer readings, not vehicle-reported odometer alone (the discrepancy between the two is often where fuel card anomalies first appear)
- R&M cost captured at invoice level per VRM, not spread by a workshop system that allocates costs to job types rather than vehicles
- Depreciation using a market-referenced residual rather than straight-line book value
- Mileage sourced from telematics, not from driver-entered figures
- Reporting cadence monthly, with rolling 3-month and 12-month views to separate one-off events from structural cost patterns
Once those inputs are in place, sorting by TCO per mile descending gives you your highest-cost vehicles immediately. The next question — whether cost is driven by a vehicle issue, a route issue, or a driver behaviour issue — requires overlaying telematics event data and route geography. But the starting point is a clean per-VRM cost number, not a fleet average that tells you roughly nothing.
The uncomfortable reality for many mid-sized fleet operators is that building this view requires integrating data from three or four systems that were never designed to talk to each other. Fuel card exports, telematics feeds, workshop records, and finance system depreciation tables all have different schemas and update frequencies. Getting them into a single per-VRM view manually is feasible but time-consuming enough that most operators do not do it consistently. That is where the cost signal stays hidden — not because the data does not exist, but because nobody has assembled it per registration mark.