TL;DR

Remote device management software does the configuration half of the job well and the physical half not at all. Everything it offers assumes the device is powered on, connected and enrolled, which is exactly the assumption that fails at the moments that cost money.

  • Everything depends on the device checking in.
  • Measure your unreachable count before buying anything.
  • Enrollment is the step that actually fails.
  • A wipe command is not evidence of a wipe.
  • Offboarding is a logistics problem, not a software one.
  • Pick the tool after you know the gap.

Every feature rests on one assumption

The device is on, connected and enrolled. Remove any one of those and the whole feature list stops.

Configuration, policy enforcement, patching, software deployment, compliance reporting, remote wipe: all of it is instructions sent to an agent that has to be listening.

That is not a flaw. It is the design, and within the assumption these tools are very good. The problem is that the assumption fails precisely when the stakes are highest: at offboarding, after a loss, during a dispute, and with anybody who has drifted out of contact.

The number that tells you where you stand

How many devices have not checked in for thirty days, as a share of the fleet. Pull it from whatever you have today, even a spreadsheet. Everything else in this article is a response to that number.

Enrollment is the step that fails

Buying is easy. Getting every machine enrolled is the part that quietly does not finish.

Three situations account for most of it. Devices already in the field before you bought the tool, which need somebody to touch them or talk their owner through it. Devices bought outside the process, on a card, by a manager in a hurry. And devices belonging to contractors, which nobody is quite sure they are allowed to enrol.

Zero touch enrollment solves the first problem for new hardware bought through the right channel, and does nothing for anything already out there. MDM for remote teams covers how that works in practice.

So a realistic plan has two tracks: new devices enrolled automatically at purchase, and a named, dated campaign to reach the existing ones. The second track is the one that gets dropped, and it is the one that determines your coverage a year later.

A command is not evidence

The gap that matters most in a regulated or contractual setting.

Issuing a remote wipe records that you issued it. Whether the device received and completed it depends on it connecting. A machine that never reconnects shows an instruction sent and nothing more.

If somebody later asks you to prove a disk was sanitised, the console entry is not the proof. NIST Special Publication 800-88 on media sanitization sets out what the standard actually requires, and the evidence contemplated there is about the completed operation rather than the request.

That is why physical recovery and certified wiping exist as a service at all, and why an offboarding process that depends solely on a console command tends to produce a paper trail that does not survive contact with an auditor.

CISA and the Secure Our World guidance is worth reading alongside, since the practices it describes sit with the organisation regardless of which tool issued the command.

Offboarding is where the model breaks

The single most predictable failure, and it is not a software failure.

Somebody leaves. Their account is disabled within minutes, which is good and is the identity layer doing its job. The laptop is at their home. It may be powered off, it may be kept deliberately, and it may simply be forgotten by everyone including them.

No remote tool resolves that. The device is not reachable, the person no longer has a reason to cooperate, and the longer it runs the less likely recovery becomes.

Four things actually help, and three of them are process rather than software. Ask for the device back before the last day rather than after it. Send packaging rather than instructions. Give it an owner with a deadline. And track it as an open item until it is physically back or written off deliberately.

our guide to IT asset management covers the register side, which is what makes the fourth item possible at all.

Choosing the software, briefly

It matters less than the two paragraphs above, which is why it comes fifth.

The useful questions are about your shape rather than features. What platforms, in what proportion. How many devices per person, since that decides whether per-user or per-device pricing favours you. How many admins. And whether you need configuration depth or just basic compliance reporting.

The category splits roughly into Apple specialists, Windows-centric suites, identity platforms with device management attached, and MSP tools that price per technician. endpoint management software covers that split.

Pricing bases differ wildly across those groups, so headline comparison is meaningless until you convert everything to your own counts.

The test that settles most shortlists

Ask each vendor what happens when a managed device stops checking in for sixty days. The answers differ more than the feature grids do, and the honest ones concede the limit quickly.

Measuring whether it is working

Four numbers, reviewed monthly, and none of them is a feature count.

Enrolled share of known devices. Not of devices the tool can see, which is circular, but of your asset register.

Checked in within thirty days, as a share of enrolled.

Devices with no assigned person. These accumulate silently and are the ones that go missing.

Open recovery items past their deadline.

Measure Measured against Tells you
Enrolled share Your asset register Coverage
Checked in within 30 days Enrolled devices Reachability
Devices with no owner Whole fleet Register hygiene
Recovery items past deadline Open items Process health

The first two measure the software. The second two measure whether your process is keeping up with it, and the second pair usually degrades first, because software keeps working while a process depends on somebody remembering to do something during a busy week.

the NIST Cybersecurity Framework is a reasonable framework to map these against if you need to present them alongside other controls. Reporting the same four every month also makes a trend visible, which a one-off audit never does, and the trend is what tells you whether the process is holding.

Closing the physical gap

Three options, and most companies end up with some mix.

Do it yourself. Workable in one or two countries, painful across many, and the cost is somebody’s time rather than an invoice, which is why it is usually underestimated.

Use local providers per country. Flexible and quickly becomes a vendor management problem of its own.

Use a specialist. RemoAsset publishes coverage of 80+ countries for procurement and 25+ for retrieval, and RemoAsset’s laptop retrieval is the retrieval side specifically.

Whichever you choose, the decision should follow the unreachable count from the first section rather than preceding it. A fleet with almost nothing unreachable does not need a third party. A fleet where a tenth has gone quiet has a custody problem that no licence addresses, and buying more software is the expensive way to avoid noticing.

What to do about devices nobody owns

Every fleet accumulates them, and they are the ones that disappear.

A machine shows in the register with no assigned person, usually because somebody left and the record was never closed, or because it was bought outside the process and attached to a team rather than a name.

Give every device a named owner, including spares, where the owner of a spare is whoever holds the cupboard key. It sounds bureaucratic and it is the single cheapest control available, because an unowned device is one nobody notices is missing until an audit.

The order that works

Measure the gap. Fix the process. Then choose the tool, sized to what is left.

Software reaches devices that check in. RemoAsset reaches the rest, retrieving laptops across 25+ countries. See RemoAsset’s laptop lifecycle management or RemoAsset’s pricing.

Frequently asked questions

What does remote device management actually cover?

The configuration half, and it rests on one assumption: the device is on, connected and enrolled. Configuration, policy enforcement, patching, software deployment, compliance reporting and remote wipe are all instructions sent to an agent that has to be listening. Within that assumption these tools are very good. The assumption fails exactly when the stakes are highest, at offboarding and after a loss.

What should I measure before buying?

How many devices have not checked in for thirty days, as a share of the fleet, pulled from whatever you have today even if that is a spreadsheet. Everything else follows from that number. A fleet with almost nothing unreachable does not need a third party; a fleet where a tenth has gone quiet has a custody problem that no licence addresses.

Why does enrollment fail?

Three situations account for most of it. Devices already in the field before you bought the tool, which need somebody to touch them or talk their owner through it. Devices bought outside the process, on a card, by a manager in a hurry. And contractor devices, which nobody is sure they are allowed to enrol. Zero touch solves the first for new hardware and nothing for what is already out there.

Does a remote wipe prove the disk was cleared?

No. Issuing a wipe records that you issued it, and whether the device received and completed it depends on it connecting. A machine that never reconnects shows an instruction sent and nothing more. If somebody later asks you to prove a disk was sanitised, the console entry is not the proof, since the recognised standard concerns the completed operation rather than the request.

Why does offboarding break the model?

Because it is not a software problem. The account is disabled within minutes, which is the identity layer working correctly, but the laptop is at someone’s home, possibly powered off, possibly kept deliberately. The device is unreachable and the person no longer has a reason to cooperate. Four things help and three are process: ask before the last day, send packaging, give it an owner, track it.

How do I choose between the tools?

By your shape rather than by features. What platforms and in what proportion. How many devices per person, since that decides whether per-user or per-device pricing favours you. How many admins. And whether you need configuration depth or basic compliance reporting. One question settles most shortlists: ask each vendor what happens when a managed device stops checking in for sixty days.

What should I track monthly?

Four numbers, none of them a feature count. Enrolled share of known devices, measured against your asset register rather than against what the tool can see, which is circular. Checked in within thirty days as a share of enrolled. Devices with no assigned person, which accumulate silently and are the ones that go missing. And open recovery items past their deadline. The last two degrade first.