What a mobile fleet actually costs, and where the money goes

Every mobile fleet business case we are shown has the same shape. Device cost, carrier plan, MDM licence, three-year term, done.

Then the programme runs, and the actual spend turns out to be somewhere else entirely. Not because anyone was dishonest — because the expensive parts of a fleet are the parts that are hard to put a number against in advance.

Here is the full picture, in the order the money usually appears.

The costs everyone counts

Devices. The visible line, and typically the one negotiated hardest. Worth noting that it is also the line where saving 8% has the least effect on the total.

Carrier plans. Predictable, and usually reviewed. Worth checking for plans still being billed for devices that left the organisation — this is a remarkably common finding.

Platform licences. MDM or UEM per device per month, plus whatever security tooling sits alongside it.

Together these are maybe half of the real cost of a managed fleet. Often less.

The costs that appear later

Staging and enrolment. Somebody has to unbox, register, enrol and configure. If devices are not registered for zero-touch enrolment before they ship, this is a person handling every device — and it recurs with every refresh and every replacement.

The support tail. Devices generate tickets: enrolment failures, forgotten passcodes, apps that will not authenticate, mail that stopped syncing after an OS update. A fleet of a thousand generates a predictable load, and it lands on someone whether or not it was budgeted.

Replacement and repair. Breakage, loss and battery degradation across a three-year term. Rarely modelled honestly at the start.

The OS release cycle. Every September, Apple ships a major iOS release, and every year some part of the environment does not survive it. Testing, remediation and the support spike are real costs that recur annually.

Certificate and token maintenance. APNs certificates, VPP and DEP tokens, and internal CA trust chains all expire. Renewing them is a small task. Missing one takes enrolment offline for the whole fleet — and that cost is measured in lost working hours across everyone affected.

Platform upgrades and vendor cases. Version currency, breaking changes, and the time spent working a vendor support case to a resolution.

The costs nobody puts in the business case

Policy drift. Policies written for a fleet that no longer exists. The cost shows up as exceptions, unmanaged devices and a compliance report nobody trusts.

The unowned platform. The single most expensive condition a fleet can be in. Nobody is accountable, so nothing is maintained, and the failures are absorbed as user frustration rather than appearing on a budget line.

Compliance evidence assembled by hand. If your Essential Eight or PSPF reporting requires somebody to build a spreadsheet each quarter, that is a recurring cost that grows with the fleet.

Secure disposal not done. Devices that leave the organisation without being wiped are not a cost until they are, at which point they are a very large one.

Lost productivity during migrations. A poorly sequenced platform migration costs far more in user downtime than the project itself. This is the argument for no-reset migration paths and staged cohorts, not just a technical preference.

How to build a number you can defend

1. Count the fleet properly first. Enrolled versus expected versus licensed. These three numbers disagree in almost every environment we assess, and the gaps are money.
2. Model per-device-per-month, all-in. Devices, plans, licences, staging, support, replacement and platform operation. One number, comparable year to year.
3. Put a cost against the expiry events. What does an enrolment outage cost you per day? That number justifies the maintenance discipline better than any argument about best practice.
4. Include the annual OS cycle. It happens every year. Budget for it once rather than absorbing it as a surprise every September.
5. Count exit and refresh. Secure disposal, data destruction certificates and the staging cost of the next generation.

The uncomfortable conclusion

For most organisations, the biggest available saving is not a cheaper device or a cheaper licence. It is moving the fleet from unowned to operated — because the unowned condition quietly generates support load, compliance risk and replacement cost that nobody attributes to it.

That is the case for a managed service, and it is also the reason we start engagements with an assessment that counts what you actually have before anybody proposes anything.