What actually belongs in cost per mile

Cost per mile is the number everyone asks for and almost nobody computes the same way. Two fleets can report figures three times apart while running identical vehicles, purely because of what they counted.

That does not make the metric useless. It makes it useful internally — as a trend on the same unit and a comparison between units of the same class — and close to meaningless as a cross-fleet benchmark.

The two categories

Split costs by whether they scale with use. This matters more than the total, because the two halves behave differently and lead to different decisions.

A high fixed share means the unit is underused — the answer is redeployment or disposal, not better driving. A high variable share means the unit is working hard, and the answer is in fuel, maintenance practice, or duty cycle. A single blended number hides which situation you are in.

The lines people get wrong

Depreciation. Purchase price less expected resale, divided by service life. Straight-line is less accurate than declining-balance early in a vehicle’s life, but it is what most fleet policies and budget offices use and it is far easier to defend. Include upfit costs in the purchase price — a plow package or a service body is a real cost that does not disappear because it was on a separate invoice.

Your own shop labour. Internal labour is a cost. Value it at a loaded rate — wages plus benefits plus shop overhead — not at the technician’s hourly wage. Fleets that only count outside repair invoices systematically understate maintenance cost, and then wonder why in-house looks so much cheaper than it is.

Overhead. Whether to allocate fleet administration into CPM is a genuine judgement call. Include it and CPM reflects the true cost of the service; exclude it and the number is cleaner for comparing units against each other. Either is fine. Doing it inconsistently between vehicle classes is not.

Accident damage. Usually exclude it, or track it separately. It reflects an event, not the cost of operating the asset, and one collision will distort a unit’s CPM for years.

Downtime. Real, expensive, and almost never in CPM — because it is hard to value and easy to argue about. Track it as a separate metric rather than trying to force a dollar figure into the cost calculation.

Why published benchmarks mislead

You will find confident figures published: industry average cost per mile of $1.82, top performers under $1.56, maintenance running $0.15–$0.18 per mile.

These circulate widely and almost all of them come from vendor content that does not disclose sample size, vehicle mix, or — critically — whether depreciation and overhead are included. Some are heavy-truck figures being quoted at light-duty fleets.

The comparisons that are actually valid:

• The same unit, this quarter against last
• Units of the same class and similar duty cycle, against each other
• Your fleet-wide figure by class, year over year

Those three will tell you everything a benchmark would have, and they are defensible in a budget meeting because you own the inputs.

Using it to make decisions

Redeployment. Units with a high fixed share and low mileage are candidates to move to a department that will use them, or to a motor pool.

Replacement timing. Rising maintenance CPM on a specific unit is one input to replacement scoring. Watch the trend, not the level — a step change matters more than a high number.

Outsourcing. Comparing your maintenance CPM against a vendor quote only works if your internal figure carries loaded labour and shop overhead. Most of the time when in-house looks dramatically cheaper, the comparison is not like-for-like.

Rate setting. For public fleets, CPM by class is the foundation of the variable component of a chargeback rate. See setting fleet chargeback rates.

Cost per hour, for equipment

For units that do significant stationary work — bucket trucks, sweepers, anything with a PTO — cost per mile is the wrong denominator entirely. A vehicle that drives 6,000 miles and runs 900 engine hours is not a low-use asset, but its CPM will say it is.

Use cost per hour for those, or track both. The general rule: if the engine runs when the wheels do not, hours are the more honest measure.