Features of a fleet management system: the counted list from 60 real products
Most lists of fleet management system features are a live map, a maintenance calendar and a fuel report, bulleted from whatever a vendor's homepage led with. They are no use for the two jobs people actually need them for: writing requirements, and judging whether the system in front of them is real.
This one is counted. We went through 60 real fleet management products across 120 official source pages, and tallied a capability only when a source directly showed it shipping. The counts tell you which features are load-bearing and which are optional, so you can scope a fleet system instead of copying a feature grid.
For each feature, this page also gives you the difference between the real version and the decorative one. Most fleet software disappointments are a feature that exists on screen but does not do the work behind it, and every entry below comes with a test you can run in a demo in a couple of minutes.
The core seven, ranked
| Feature | Products (of 60) | Share |
|---|---|---|
| Fuel management and consumption | 41 | 68% |
| Dashboards and reports | 41 | 68% |
| Vehicle and asset register | 40 | 67% |
| Preventive maintenance schedules | 40 | 67% |
| Driver app and field capture | 32 | 53% |
| Digital inspections and checklists | 31 | 52% |
| Expenses and cost per mile | 31 | 52% |
Integrations and APIs also appear in 39 of 60, but they are deliberately kept off the core list: integration plumbing is a branch you choose, and it must never become the organising centre of a fleet product. The same goes for the live map, which we come to below.
1 · Fuel management and consumption — 41 of 60
The most evidenced capability in the category, and the one most often faked.
Decorative version: a fuel log with a date and an amount spent. Real version: every fill records the quantity, the odometer reading and the price, and consumption is calculated from the distance between fills.
The test: record two full fills with odometer readings and work out the consumption by hand. The system should show the same number. Then record a fill without an odometer reading. It should refuse or flag it, not quietly produce a figure. A fuel entry without an odometer reading cannot produce a consumption figure at all, so any system that shows one anyway is estimating.
2 · Dashboards and reports — 41 of 60
Fleet status, costs, overdue work, utilisation.
Decorative version: tiles with numbers on them. Real version: every figure opens the rows behind it.
The test: click a tile — "Open defects: 5" — and count what it opens. If it opens four, or nothing, the tile is decoration. A dashboard that cannot be reconciled with the records it summarises is worse than no dashboard, because people make decisions from it.
3 · Vehicle and asset register — 40 of 60
The product centre. Everything else in a fleet system is something that happens to a vehicle.
Decorative version: a table of vehicles with a registration and a model. Real version: each vehicle opens a record carrying its status, its driver, its inspections, its maintenance and its fuel history.
The test: open one vehicle. From that one page, can you see its last inspection, any open defects, its next service and its recent fuel? If you have to go to four other screens and filter by registration, the register is a list, not a record.
4 · Preventive maintenance schedules — 40 of 60
Decorative version: a to-do list of services with due dates. Real version: intervals in both distance and time, with a next-due that recalculates when the work is completed.
The test: mark a service complete and look at the next due date. It should move forward by the interval, in both miles and days, from when the work was actually done. If it does not move, or only one of the two moves, the schedule never advances on its own.
5 · Driver app and field capture — 32 of 60
Decorative version: the desktop form, squeezed onto a phone. Real version: short, single-task forms built for fast numeric entry, because they are used at a pump, in a yard, with gloves on.
The test: record a fuel fill on a phone, one-handed. If it takes longer than half a minute or needs scrolling, drivers will not do it, and every figure downstream depends on their readings.
6 · Digital inspections and checklists — 31 of 60
Decorative version: a checklist that saves when you submit it. Real version: a failed safety item raises a defect linked to the vehicle and takes the vehicle out of service. The inspection is a gate, not a form.
The test: fail a safety item on a pre-trip check, then try to assign that vehicle to a driver. It should be refused until the defect is resolved. If the vehicle goes out anyway, the checklist is a record of a problem that nobody acted on.
7 · Expenses and cost per mile — 31 of 60
Decorative version: a cost-per-mile figure on the dashboard. Real version: total spend for a period divided by the distance driven in that period.
The test: check one vehicle's cost per mile against a hand calculation. If it comes out at a fraction of a cent, the system is dividing by the odometer readings instead of the distance driven — an error of roughly two orders of magnitude, and one we found in a real build.
The features marketing undercounts
Below the core seven sits the part of a fleet system that holds it together, and it is barely advertised:
| Feature | Products (of 60) |
|---|---|
| Service and repair history | 20 |
| Work orders and repair execution | 18 |
| Defect, fault and issue management | 14 |
| Driver and operator records | 13 |
| Roles, permissions and scoped access | 11 |
| Audit trail and operating history | 8 |
These are low because product pages lead with vehicles and maps, not because fleets do without them. Read them as one chain: an inspection fails, which raises a defect; the defect becomes a work order; the work order returns the vehicle to service and writes the repair history. The defect is the link — only 14 of 60 products document it, yet nothing connects without it.
The same is true of people. A vehicle without an accountable driver cannot say who inspected it or who had it when something happened. And a driver who can read the whole fleet's records, rather than their own vehicle's, is a defect whatever the market documents. A system that skips these is a collection of screens that happen to share a vehicle list.
What is deliberately not on the list
Several families are common but belong to particular fleets rather than to every fleet:
- Live GPS map — 30 of 60. Common, not universal. A maintenance-and-inspection system can be complete without one, and a map that shows last-known position must say so rather than implying a live feed.
- Regulated compliance — 31 of 60. Electronic logging, hours of service, fuel-tax reporting, tachographs, driver vehicle inspection reports. These are chosen by jurisdiction, never grouped behind a vague "Compliance" page.
- Alerts and exception queues — 22 of 60. Where a fleet manager starts the day, but only useful if they reconcile with the records they link to.
- Dispatch, jobs and routes — 21 of 60. A separate job queue of who, where and when, kept apart from maintenance because the people who use each need different tools.
- Parts inventory and purchasing — 14 of 60. Real only when quantities reconcile across ordering, receiving, issuing and returns.
- Driver licence and eligibility — 5 of 60. Required as soon as you operate regulated vehicles.
- Offline capture — 3 of 60. An explicit decision for fleets in poor coverage, because offline sync that does not really work is worse than none.
Each one is right for somebody and drags its own data model in with it. Pick them deliberately.
The requirements list you can hand someone
Written to be checkable, because a requirement nobody can test is a preference.
- A vehicle record that shows status, driver, inspections, defects, maintenance and fuel on one page.
- Every fuel entry stores quantity, odometer and price. Consumption is calculated, never estimated, and an entry without an odometer is refused or flagged.
- Maintenance intervals in both distance and time. Next-due recalculates from the completion date.
- A failed safety item on an inspection raises a linked defect and blocks the vehicle from assignment until it is resolved.
- Defects become work orders, and closing a work order returns the vehicle to service and writes the history.
- Cost per mile is period spend divided by distance driven, checked against a hand calculation.
- Every dashboard figure opens the records it counts.
- Drivers see their own vehicle and their own records. Managers see the fleet.
- A fast, single-task mobile form for fuel, odometer and inspections.
- Branches chosen in writing: [map / compliance by jurisdiction / dispatch / parts / offline].
Why AI-built fleet systems miss these
Ask a coding agent for fleet management software and it will build the live map first, because it is the most photographed screen in the category. Then a vehicle list, a maintenance calendar and a dashboard, and they will all look right. What it will not build unprompted is the behaviour: fuel that stores the numbers consumption needs, an inspection that stops a vehicle, a dashboard that adds up.
When four AI models built the same fleet specification, the gaps were exactly these. The lowest-scoring build accepted a fuel reading and threw away the quantity and the odometer, and stored costs with no amount at all. Another wrote a fixed fuel-efficiency figure under a heading that said it was calculated. All four builds are live — you can sign in and compare them on the fleet management demo page.
The difference between them was not the model. It was how much of the list above the specification made non-optional.
If you are still running the fleet from a spreadsheet, fleet management in Excel sets one up properly and shows where it stops, and free fleet management software covers what free plans hold back. For the build order, see how to build a fleet management system.
If you are pointing a coding agent at this, the Fleet Management planner is the specification we hand ours: this inventory, the inspection-to-repair chain made explicit, and the branches as deliberate choices.