Choosing a Retrieval Partner When Devices Sit in Twelve Countries

TL;DR

• Retrieval coverage is always narrower than delivery coverage, because getting a device back from a private address across a border is harder than sending one out.
• A vendor can genuinely ship to a country and have no practical route for recovering a machine from it. Ask about collection as a separate question.
• Packaging decides the return rate more than anything else. A prepaid, pre-labelled box at the door beats instructions by a wide margin.
• The trigger is the real weak link. Nothing automatically connects a resignation to a collection, so most chasing starts after the person has stopped answering.
• Nobody in this category publishes a price. Six vendors checked on their own pages on 14 October 2026 and not one shows a figure.
• Decide in advance which devices are not worth recovering. The rule costs nothing and saves three months of chasing to reach the same conclusion.

Coverage Maps Are Drawn for Delivery

Every vendor in this category publishes a coverage number and every one of them is counting outbound shipments. That is the easy direction: a courier serves the destination, a parcel goes out, somebody signs for it.

Recovery runs the other way and almost everything about it is harder. The sender is a former employee with no obligation to you. The pickup is from a residential address, often during working hours when nobody is home. The parcel crosses a border in the direction that customs authorities scrutinise more closely. And the whole thing depends on somebody who has already moved on bothering to participate.

So a vendor's retrieval footprint is a different and shorter list than its delivery footprint, and almost nobody publishes the second one. Across twelve countries, the gap between what a vendor can deliver to and what it can genuinely collect from is where your unreturned devices come from.

Recommended Tools

Best tools for Device Management

B BambooHR HR management system L Lattice Performance management D Deel Global payroll & EOR

View all recommended tools

The Trigger Problem Comes First

Before vendor selection, the thing that decides your return rate is whether anything starts the process in time.

At most companies nothing does. A resignation lives in an HR system that knows nothing about hardware. So the collection begins when somebody remembers, which is typically after the last working day, which is precisely the point at which the former employee stops reading email from you.

The fix is a trigger at notice rather than at exit. Whoever processes a resignation adds one line: tell the person who handles devices, with the leaver's name, country and last working day. That is it, and it moves the whole exercise several weeks earlier into the period when the person is still employed, still reachable and still motivated to be helpful.

Book the collection while they are still working. A courier slot agreed in their final week has a dramatically better success rate than one arranged afterwards, because they are expecting it and because they still think of themselves as somebody who does what the company asks.

No vendor fixes this for you. A retrieval partner with excellent coverage and a trigger that fires three weeks late will still lose devices, and that failure will look like the vendor's rather than yours.

What Actually Determines the Return Rate

Packaging, by a distance. A former employee asked to find a suitable box and arrange a courier mostly does not, and it is not malice. It is an errand for an organisation they no longer work for. A prepaid, pre-labelled container arriving at their door removes every decision from them except handing it over, and the difference this makes is larger than any other single factor.

How many actions you require. Each step is an opportunity to stop. Receive packaging, pack the device, book a slot, be home, hand it over. Vendors who compress that, by sending packaging with a pre-booked collection date, do better than those who send a link to a scheduling page.

Whether somebody is chasing, and who. A polite sequence from a named person works better than automated reminders from a system. It is also work, and it is worth knowing whether the vendor does it or whether it reverts to you.

And the condition record. A machine arriving damaged is an argument you cannot win without a dispatch condition note and a photograph. Ask whether the vendor records condition at both ends and whether you can see it.

The Countries That Will Be Difficult

Not all twelve are equal, and the hard ones are predictable enough to plan around.

Markets where the vendor has no presence are served by a courier booking and nothing else. Collection is possible and slow, and nobody is locally accountable if it stalls.

Countries with strict import controls make the return journey genuinely complicated, because a used device crossing a border is treated differently from a new one and may need documentation nobody has prepared.

Anywhere with a small headcount. One person in a country means no local arrangement is worth building, so every collection is a one-off.

And anywhere you are hiring through a third party, because the relationship that ends may not be one you control, and the notice may reach you late or not at all.

Ask every vendor to mark your twelve countries as collection with local presence, collection via partner, or outbound only. That single exercise takes them ten minutes and tells you where your losses will come from.

The Thirty-Four Machines

A 400-person company operating across twelve markets ran its offboarding the way most companies do, which is to say nobody ran it. Leavers were processed by People, devices were a line on a checklist, and somebody in operations sent an email asking for the laptop back when they noticed.

An audit eighteen months in found thirty-four devices with people who had left. The oldest had been gone over two years. Several were in markets where nobody had ever attempted a collection, because the first attempt had been an email that went unanswered and no second attempt existed.

The replacement cost of thirty-four machines is a real number. What got attention internally was something else: thirty-four laptops holding company data, outside any control, with people who had moved on to other employers. Nobody had thought of unreturned hardware as a data question until somebody put it that way in a meeting.

The fix was not a vendor. It was a trigger at notice, prepaid packaging, a named chaser, and a written rule about what was not worth recovering. The vendor conversation came afterwards and was much easier, because by then the company knew what it was asking for.

What Good Looks Like, Step by Step

It helps to see the whole sequence, because each step is where a different share of devices gets lost.

Notice received. Somebody tells the device owner the same day, with name, country and last working day. This single step recovers more machines than any vendor choice.

Collection booked while they are still employed. A date in their final week, agreed with them directly, while they still answer email and still feel some obligation.

Packaging arrives before the last day. Prepaid, pre-labelled, with the collection already scheduled. No decisions left for them to make.

Device condition recorded at handover. A photograph and a note, ideally from them, which settles any later dispute about damage.

Collection happens, tracked. Somebody sees that it did. A booking is not a collection, and the gap between the two is where the process quietly fails.

Receipt, assessment and wipe. The machine arrives somewhere, is checked against the condition record, wiped, and either stored or reissued.

Register updated. The row closes, with the date and the outcome.

And an escalation path for the ones that stall. A second attempt, a third, and a stated point at which the device is written off rather than chased indefinitely.

Eight steps. Most companies have two of them, and the two they have are the first and the last.

The Options

RemoAsset, named below, is owned by the same people who publish HROpsLab. Its limitations are stated as plainly as everybody else's.

RemoAsset

Procurement, deployment, tracking and retrieval for distributed teams, with devices enrolled into your management system before dispatch.

On retrieval specifically: it holds the serial, the assignment and the original delivery address from its own fulfilment, which makes a collection easier to initiate for devices it supplied. That is also its limit in this context: it is noticeably weaker for hardware it did not supply, so a fleet bought elsewhere is exactly the case to test. It also publishes no price and requires a demo, and is not a device management system.

Workwize

Hardware lifecycle with owned warehouses. Domain is now goworkwize.com; the older address no longer resolves, verified 14 October 2026.

On retrieval: warehousing matters here more than almost anywhere, because a recovered device has to go somewhere and be checked, wiped and either stored or redeployed. Vendors without warehouses ship it to your office or to a partner. No published price.

Firstbase

Equipment on a subscription basis with retrieval included by design.

On retrieval: included rather than priced as a module, which is a genuine structural advantage at a high leaver rate, since per-collection billing compounds in exactly the quarters you least want it to. No published price.

GroWrk and allwhere

Both scope arrangements per customer, GroWrk across a wide country list and allwhere closer to a managed service.

On retrieval: a scoped arrangement can be written to cover the specific difficult countries rather than offering an even service everywhere, which suits a twelve-country footprint with three hard markets. Neither publishes a figure.

Deel IT

Devices attached to an employment platform.

On retrieval: the trigger advantage is real, because the provider processes the termination and therefore knows first. Collection coverage is a separate question and worth asking per country. No published price.

The Comparison

The middle column matters more than people expect. A recovered laptop has to be received, assessed, wiped and either stored or reissued, and a vendor with nowhere to put it will send it to you.

The Decision Table

Deciding What Not to Recover

The decision nobody writes down, and it saves more than any vendor choice.

A four-year-old laptop with somebody in a distant market frequently costs more to collect than it is worth. Packaging, courier, customs, somebody's time chasing, and the assessment at the other end can exceed the residual value of the machine comfortably.

Write the rule in advance. Something like: devices over a stated age in countries outside your core markets are written off rather than recovered, with the value recorded and the register row closed. Two sentences.

Doing it in advance is the whole point. Without a rule, every device gets chased, and the ones not worth chasing get chased for three months before somebody reaches the same conclusion with three months of effort already spent.

One exception worth keeping. Data, not value. A machine holding sensitive material may be worth recovering regardless of residual, or may need a remote wipe instead, which is a reason to have a management system rather than a reason to collect. Decide that separately from the economics.

And record the write-off properly rather than deleting the row. A device written off should stay in the register marked as such, with its last known holder, the date and the value. Deleting it loses the evidence that the machine ever existed, which matters if it surfaces later or if somebody asks how many devices the company has lost over three years. A write-off somebody decided is a different thing entirely from a device that quietly vanished.

Measuring Whether It Works

Retrieval is one of the few things in this area with a clean number attached, and almost nobody tracks it.

Return rate, as a percentage of leavers. Devices recovered divided by devices issued to people who left, over a rolling twelve months. One figure, and it tells you more about the arrangement than any service level does.

Days from last working day to receipt. The median rather than the average, because a handful of two-year stragglers distort the mean beyond usefulness. A median under three weeks is good. Over six suggests the trigger is firing late.

Outstanding over ninety days. A count rather than a rate, because each one is a specific laptop with a specific former employee and the list should be short enough to read.

And the rate by country. This is where a vendor's real coverage shows up, and it is the evidence you bring to a renewal conversation. Two markets with materially worse outcomes than the rest is either a vendor problem or a local process problem, and the distinction is worth establishing before blaming either.

Pull these quarterly, not monthly, because retrieval has a long tail and a monthly figure is mostly noise. Four data points a year is enough to see a trend and few enough that somebody will actually produce them.

What a Vendor Cannot Fix

Being fair about the limits, because it is easy to read an article like this as suggesting the right partner solves everything.

They cannot create a trigger inside your HR process. The notification has to come from you, and a vendor with perfect coverage waiting for an instruction that arrives three weeks late will perform badly through no fault of their own.

They cannot make a former employee cooperate. Good packaging and good chasing improve the odds substantially and nothing guarantees it, and a small residual of people simply will not engage regardless of what anybody does.

They cannot recover what they do not know about. A device absent from the register is a device nobody will chase, which is why the register and the retrieval arrangement are the same problem viewed from two angles.

And they cannot decide what is worth recovering. That is a commercial judgement about your own hardware, and leaving it to the vendor means either everything gets chased or nothing does.

So weight the vendor choice accordingly. It decides collection coverage, packaging quality and whether somebody chases. It does not decide the four items above, and those are collectively responsible for more lost devices than coverage is.

What Getting This Wrong Costs

The obvious cost is the devices. At a normal leaver rate across twelve countries, a poor retrieval arrangement loses a meaningful number of machines a year and each one is a full replacement cost.

The less obvious cost is the register. Every unrecovered device stays on your asset list as something you theoretically own, or gets quietly deleted, and both are wrong. The first inflates what you think you have and the second loses the record that a machine existed at all. Neither is discovered until somebody counts.

And there is the data exposure, which is the one that matters most and gets the least attention. A laptop sitting with a former employee is company data outside your control, and the length of time it sits there is a direct function of how quickly your retrieval process starts. That is a security question wearing a logistics costume, and it is the strongest argument for fixing the trigger rather than the vendor.

Ask it this way. How many devices are with people who left more than ninety days ago, and who is chasing each one?

Where Recovered Devices Should Go

The question that gets asked last and shapes the arrangement more than people expect: what happens to a laptop after it comes back.

Straight to another employee is the best outcome and the hardest to arrange, because it requires a joiner who needs that specification at roughly the moment a leaver returns one. Without warehousing and a wipe-and-check step in between, this almost never lines up.

Into storage is the realistic answer most of the time. Somebody holds it for weeks or months until it is needed. Vendors with warehouses do this natively and charge for it, which is honest. Vendors without send it somewhere, and that somewhere is frequently your office.

To your office is the default when nobody planned. It works at small scale and becomes a cupboard full of machines nobody has assessed, wiped or recorded, which is how a recovered device turns into a lost one by a slower route.

Or to disposal, for anything at the end of its life, which needs a proper certified route rather than a bin, and that is a separate vendor decision with its own requirements.

So ask early where recovered machines physically go under the arrangement you are buying, and what each option costs. A retrieval service that returns devices to you successfully and leaves you with a storage problem has solved half the job and handed back the other half.

Contractors and Third-Party Staff

The population most likely to keep a device, for a structural reason rather than a behavioural one.

No leaver event fires. An engagement ending is not an event anybody experiences. Work tapers, a final invoice arrives at some point, and the relationship fades rather than ending on a date. So nothing triggers, because the thing that normally triggers did not happen.

The relationship may not be yours. Where somebody works through an agency or a studio, your contract is with the firm and the device is with a named individual who may rotate off without anybody telling you.

And the budget holder is often the only person who knows. They are also not thinking about hardware, so the notification has to be asked for rather than expected.

Three fixes, none of which involves a vendor. Record the contract end date at the start, because it exists on day one and is the only reliable trigger available. Put a review thirty days before it with a named owner. And where the relationship is with a firm rather than a person, make named changes a condition of the arrangement, so somebody tells you when an individual joins or leaves the engagement.

Then recover exactly as you would from an employee. The mechanics are identical once a trigger exists; it is only the trigger that is missing.

When You're Ready to Choose a Partner

Three things to produce first, and they take an afternoon.

Your leaver list by country for the last twelve months, with how long each collection took and which ones never completed. Most companies have never assembled this and it is the single most useful document in the evaluation, because it tells you and the vendor exactly where the problem is.

Your twelve countries ranked by headcount, so vendors classify the ones that matter rather than the ones that sound interesting.

And the not-worth-recovering rule, written, so the arrangement you buy covers the devices you actually intend to chase rather than all of them.

Then ask every vendor the same three questions: which of these countries can you collect from and by what route, what is your typical return rate for home collections, and who chases. The second question separates vendors more than anything else and almost nobody asks it, partly because a vendor who answers honestly looks worse than one who does not answer at all.

A good answer to that second question includes a caveat, the same way a good answer about replacement does. Something like a high rate in markets with a local presence, noticeably lower in two named countries, with a suggestion about how to handle those. A vendor volunteering where they are weak before you discover it is the most reliable signal available in this whole category, and it is worth more than a confident number with nothing behind it.

And bring the leaver list to the first call rather than the third. A vendor looking at your actual countries, your actual volumes and your actual failures will quote and advise against reality, which is both more useful and usually more expensive than the version produced from a tidy summary.

Frequently Asked Questions

HROpsLab takes no vendor money and publishes no paid placements.