TL;DR
- Device as a service bundles four things into one monthly figure: the hardware, getting it to the person, supporting it while they have it, and taking it away at the end. You pay for all four whether or not you need all four.
- If you buy 20 laptops every couple of years, hand them out in an office, and have somebody who can swap a broken one, you do not need this. You need a purchase cycle.
- The value is almost never the hardware. It is the deployment and the end-of-life handling, which are the two parts most companies do badly and rarely cost properly.
- The market splits into manufacturer programmes, remote-lifecycle platforms, and financing arrangements with services bolted on. They look alike in a brochure and behave very differently at month 37.
- Nobody in this category publishes a price. Every number comes from a sales conversation, so comparison happens on structure rather than on cost.
- Done properly, a new starter has a working machine on day one in a country you have never shipped to, and you never see the laptop again after they leave.
The Contract That Was Cheaper Until Month Thirty-Seven
A 180-person company moved to a device subscription in its second year of remote hiring. The arithmetic was persuasive. No capital outlay, a predictable monthly figure per head, machines refreshed on a three-year cycle, and somebody else handling the shipping to twelve countries. For two years it was exactly what was promised.
Then the contract reached its final quarter and three things arrived at once. Devices that staff wanted to keep using were due back. Machines that had been damaged carried charges nobody had modelled. And the renewal quote assumed a headcount the company had already passed, so the per-unit figure moved. None of that was hidden. All of it was in the agreement. Nobody had read the end of the contract because the beginning of it had solved a real problem.
The real issue is not that subscriptions are a bad deal. For a distributed team they are frequently the only practical deal. It is that device as a service is a financing decision wearing an IT decision's clothes, and the parts that matter most sit at the end of the term rather than the beginning. The monthly figure is what gets compared. The return condition, the refresh trigger and the exit are what determine whether the arrangement was good.
Best tools for Device Management
This is what device as a service is supposed to solve, and it does solve it, provided you know which of the four bundled things you are actually buying.
When You Don't Actually Need Device as a Service
When the manual way is genuinely fine
One or two offices, predictable hiring, and somebody on site who can hand over a laptop and keep a spare in a cupboard. At that shape, buying machines outright is cheaper and simpler, and the services in a subscription duplicate what you already have. The honest advice is a refresh budget and a spares shelf.
When friction starts appearing
The signal is a new starter waiting. Not once, which is bad luck, but a second time, which is a process. When the gap between an accepted offer and a working machine is measured in weeks because somebody is researching couriers, the deployment half of a subscription starts earning its cost.
When it becomes a liability
Capital constraints turn this from convenience into structure. A company that cannot spend 90,000 pounds on hardware in a quarter but can spend a monthly figure is not choosing a service model for operational reasons, it is choosing it for balance sheet reasons, and that is a legitimate and common driver. It becomes a liability question when the alternative is people working on unsuitable machines.
The edge case that forces it
Hiring into a country where you have no entity, no office and no relationship with a supplier. Buying hardware locally means an entity or an expense claim, importing it means customs, and shipping it from headquarters means a two-week delay and a duty bill. This is the case where a subscription is not an optimisation but the only route that works.
Five Questions People Ask at the Point of Signing
"What exactly is in the monthly figure?" Four things, in varying proportions: the hardware, deployment, support, and end-of-life. Vendors weight them differently and rarely break them out.
"What happens at the end of the term?" The most important question and the one asked last. Return, extend, or buy out, and the terms for each differ enormously.
"Who decides a device is damaged?" Whoever inspects it, under a standard written into the agreement. Get the standard, in writing, before signing.
"What if headcount drops?" Most agreements assume growth. Find the clause covering reduction, because it is the one that bites in a downturn.
"Is this cheaper than buying?" Usually not, on hardware alone. It is frequently cheaper once deployment and recovery are costed honestly, which is the comparison almost nobody runs.
What the Monthly Figure Is Actually Made Of
Breaking the bundle apart is the single most useful thing you can do before a sales call, because it tells you which parts you are paying for and would otherwise do yourself.
The hardware. A laptop with a known cost and a known resale value. This part you could buy. On its own, financing it through a subscription is more expensive than purchasing it, which is how financing works and is not a scandal.
Deployment. Getting a configured machine to a person in a country, through customs, on a date. This is the part companies consistently underestimate, because the cost lives in somebody's week rather than on an invoice. If your answer to "how does a laptop reach a new hire in Portugal" involves a person researching couriers, you are already paying for deployment, just not visibly.
Support. Replacement when a machine fails, and a route to a working device that is not "post it to head office". For a distributed team with no local IT, this is worth more than it looks and is usually the second-largest real component.
End of life. Collection, data destruction, and either resale or disposal. The part that is invisible on day one and decisive at month 37. A subscription that takes the device back has removed a problem most companies solve badly or not at all.
| Component | Could you do it yourself | What it costs you today, unpriced |
|---|---|---|
| Hardware | Yes, and more cheaply | Capital, and a refresh decision every 3 years |
| Deployment | In your own country, yes | Somebody's week per international hire |
| Support | Only where you have cover | Days of lost productivity per failure |
| End of life | Rarely well | Unrecovered devices and no destruction evidence |
The pattern that emerges for most distributed teams is that they are buying deployment and end-of-life, and accepting a worse price on hardware in exchange. That is a sound trade if you say it out loud. It is a poor trade if you believed you were getting cheap laptops.
What "Support" Means When Nobody Is Nearby
Support is the component buyers scrutinise least and use most, partly because the word is familiar and everybody assumes they know what it covers. In a distributed context it means one specific thing: how quickly somebody whose laptop has died gets a working one, where they are.
Three models hide behind the same word, and the difference is days.
Advance replacement from regional stock. A machine ships from a warehouse near the person, usually before the broken one comes back. This is the only version that produces a next-day answer in most countries, and it depends entirely on whether the provider holds stock in the region rather than on how good their service desk is.
Repair and return. The device goes away, gets fixed, and comes back. Reasonable for a non-urgent fault, unworkable for a dead machine, and the elapsed time is frequently 2 to 3 weeks once international shipping is involved in both directions.
Ship from a central hub. A replacement leaves a single warehouse and crosses a border, with the customs delay and the duty exposure that implies. Better than nothing, and rarely better than a week.
So the question to ask is not whether support is included. It is where the replacement physically comes from for each of your main markets, and whether stock is held there. A provider with excellent service levels and no regional stock will still take 10 days to get a laptop to somebody in Chile.
| Support model | Realistic time to working device | Depends on |
|---|---|---|
| Advance replacement, regional stock | 1 to 3 days | Warehouse presence near the user |
| Ship from central hub | 5 to 10 days | Customs, carrier, duty handling |
| Repair and return | 2 to 3 weeks | Shipping in both directions |
And ask what happens to the broken one, because the answer reveals whether end-of-life is genuinely handled or quietly left with you.
The Three Shapes This Comes In
Manufacturer programmes
What it is. The hardware maker offers its own fleet on subscription, with its own services wrapped around it. Lenovo's TruScale is the most frequently cited, and the other major manufacturers run equivalents.
When it's right. Large, standardised fleets in countries the manufacturer serves directly, where a single hardware brand across the company is acceptable or desirable.
When it fails. Flexibility. You are on one manufacturer's catalogue, which is a problem the first time a team insists on a different platform. Coverage in smaller markets is thinner than the brand suggests. And worth knowing: the TruScale product URL recorded in our own data now returns a 404, although the brand is still named on Lenovo's services pages, so check where the current programme actually lives before quoting anything from an older comparison.
Remote lifecycle platforms
What it is. Software-led providers built around distributed teams, handling procurement, delivery, support and retrieval across borders.
When it's right. When your constraint is geography rather than volume. Fifty people in eleven countries is the shape these are built for, and the manufacturer programmes handle it poorly.
When it fails. Unit economics at genuine scale. At several thousand devices in a handful of countries, a manufacturer programme or a direct lease will usually beat them on price, because you are paying a premium for exactly the flexibility you no longer need.
Financing with services attached
What it is. A lease or similar instrument, with deployment and support added by the lessor or a partner.
When it's right. When the driver is genuinely the balance sheet, and the services are secondary. It is also the most transparent of the three about what the money is doing.
When it fails. The services are often thinner than a specialist's, because the core business is finance. If deployment to twelve countries is your actual problem, this shape solves the wrong half. The financing comparison itself, lease against buy, is a separate question with its own arithmetic and deserves its own analysis rather than being decided inside a device conversation.
Reading the Contract's Last Page First
Most of the risk in these agreements sits in three clauses, and all three are at the end.
Return condition. Define what fair wear and tear means, and who decides. The gap between "minor cosmetic marks acceptable" and an unqualified requirement that devices be returned in good condition is the gap between a predictable end of term and a surprise invoice. Ask for the actual inspection standard and an example of a charge.
The refresh trigger. Whether the clock runs from contract start or from each device's deployment date. For a company hiring continuously these produce very different outcomes, because a device issued in month 30 of a 36-month term is barely used when it is due back.
Headcount movement in both directions. Growth is always provided for. Find the clause for reduction, and read it assuming a bad quarter rather than a good one. Minimum commitments are the normal shape and they are not unreasonable, but they should be a decision rather than a discovery.
What to do:
- Ask for the return condition standard in writing, with an example of a chargeable fault.
- Establish whether the term runs per contract or per device.
- Find the clause covering a 20 per cent headcount reduction and read it twice.
- Ask what a buyout costs at month 24 and at end of term, as figures.
- Confirm who holds the data destruction obligation and what evidence you receive.
- Confirm which countries are served directly and which are subcontracted.
Running the Comparison Nobody Runs
The article has twice referred to a comparison most companies skip. Here it is, as a method you can complete in an afternoon with figures you already hold. The point is not to produce a precise number. It is to stop comparing a monthly subscription figure against the price of a laptop, which is the comparison that makes every subscription look expensive and is the wrong one.
Take a three-year window and build both sides.
The buying side starts with the obvious and then keeps going. Purchase price per machine, multiplied by the number you would buy over three years including replacements for failures and for growth. Then add the four things that never make it onto the spreadsheet.
Coordination time per international hire. Count the hours somebody spends arranging a machine for a starter in another country, from choosing a supplier to confirming delivery, and multiply by the number of international hires you expect. For most companies this is the largest hidden figure and it is carried by a person whose job description says something else entirely.
Productivity lost to failures without local cover. Take your failure rate, which your ticket history knows, and multiply by the realistic days to get a working replacement into somebody's hands where they are. A machine that dies in a city with no office is not a four-hour problem.
Devices that never came back. You calculated this if you have read the asset recovery side of this topic; if not, pull last year's leavers and count. Each one is a full purchase price written off.
Devices recovered and never reissued. The laptops in the cupboard. These cost the same as the ones that never came back, because the money was spent and no value was returned, and they are more irritating because you did the hard part.
The subscription side is simpler and you should still pressure it. Monthly figure times headcount times 36, plus whatever the end-of-term clauses imply. Ask the vendor for an end-of-term estimate in writing, including a realistic allowance for devices returned with damage, because that line is routinely omitted from proposals and routinely appears on invoices.
| Line | Buying | Subscription |
|---|---|---|
| Hardware over 3 years | Purchase price plus replacements | Inside the monthly figure |
| Getting devices to people | Hours per international hire | Inside the monthly figure |
| Replacement when one fails | Days of lost work | Inside the monthly figure |
| Recovery at the end | Usually unpriced, often unmanaged | Inside the monthly figure |
| Devices written off | Full purchase price each | Return condition charges |
| End of term | You still own ageing hardware | Return, extend or buy out |
Two things usually fall out of this. The first is that buying wins clearly for a concentrated domestic fleet, and no amount of vendor framing changes that. The second is that for a genuinely distributed team the two sides land closer together than anybody expected, and the decision stops being about money and becomes about which set of problems you would rather own.
And if the exercise produces a number you do not believe, the useful response is not to adjust it. It is to notice which line you could not fill in, because that line is a cost you are carrying without visibility, and finding it was worth more than the comparison.
Measuring Time to a Working Laptop
The other number the evaluation turns on, and it takes an hour to produce.
Pull the last 10 starters. For each, find two dates: the day the offer was accepted and the day they had a configured machine they could work on. Not the day it shipped, and not the day it arrived in the country. The day they were working on it.
Then split the list in two, domestic and international, and take the median of each. The gap between those two medians is the thing a subscription is selling you, and it is usually much wider than people's impression of it, because the domestic cases are frequent and memorable while the international ones are occasional and quietly awful.
What to do:
- Record both dates for the last 10 starters, separated by country.
- Take the median of each group, not the average, so one disastrous case does not distort it.
- Note how many of the international cases involved somebody improvising a solution.
- Ask the vendor for their committed delivery time into your three most common hiring markets.
- Compare their commitment against your actual median, not against your target.
So if your domestic median is 2 days and your international median is 16, you have the case already and the rest of the evaluation is about which provider covers your markets. But if both medians are tight, the deployment half of the bundle is worth very little to you, and you should be buying this for end-of-life or not at all.
How to Choose: Five Questions Before You Talk to Any Vendor
What is a new hire's time to a working laptop today? Measure it for the last ten starters, separating domestic from international. That number is the thing a subscription is actually selling you, and if it is already three days you are buying less than you think.
What did you fail to recover last year? Count the devices that did not come back and the ones that came back and were never reissued. Both are end-of-life costs you are already paying invisibly, and together they are usually the strongest part of the case.
Which countries are genuinely in scope? Not the vendor's coverage map, your hiring plan. Ask specifically which markets are served directly and which through a partner, because the answer determines delivery times more than anything else in the proposal.
Is the driver operational or financial? If it is the balance sheet, say so internally, because it changes which of the three shapes fits and it changes who needs to approve. Conflating the two is how companies buy a service model to solve a cash problem and then judge it on service.
Who owns the relationship in year two? These agreements need somebody who reads the renewal before it arrives. Without that, the end-of-term clauses become whatever the vendor's template says, and the company discovers them in the quarter they apply.
Seven Options Worth Knowing
Every vendor below was checked against its own pricing page on 6 and 7 October 2026. None publishes a figure. Two describe a model without numbers: GroWrk names a per-order option and a subscription tier, allwhere refers to pay-as-you-go and fixed rates. Budget for a sales process, because there is no route around it.
RemoAsset
Disclosure: RemoAsset is owned by the same people who publish HROpsLab. It appears here because it competes in this category and is assessed on the same criteria, with its limitations stated in the same detail as every other option.
Best for: distributed teams that want one platform holding the whole lifecycle rather than a subscription for hardware and separate arrangements for everything else.
Why companies choose it: procurement, delivery, storage, retrieval and wiping run from one record, so the end-of-life half that usually gets neglected is handled by the same system that issued the device. Offboarding triggers the return automatically. For teams whose pain is geography rather than volume, that closes the loop that otherwise leaks devices.
Where it struggles: it publishes no price and requires a demo. It is a lifecycle platform rather than a financing instrument, so if your actual driver is converting capital expenditure into operating expenditure, that is a conversation to have explicitly rather than assume. It is not an MDM, so device configuration and policy enforcement need something else, and it is not a certified disposition vendor, so bulk end-of-life with disposal certification is a separate arrangement.
Workwize
Best for: multi-region fleets where storing devices near people is the binding constraint.
Why companies choose it: regional warehousing, which is what makes reissue practical rather than theoretical, with strong European coverage.
Where it struggles: no published price, and the multi-region machinery is more than a company operating in three countries needs.
Deel IT
Best for: companies already running Deel for employment or contractors.
Why companies choose it: equipment and employment in one relationship, so the hire and leave signals already exist without an integration.
Where it struggles: quote-based, and much more compelling as an extension of an existing Deel footprint than as a standalone device decision.
Firstbase
Best for: companies providing a whole home setup rather than a laptop.
Why companies choose it: furniture and peripherals are in scope alongside computers, which matches what many companies actually issued.
Where it struggles: no published price, and the breadth is only worth paying for if your policy genuinely covers desks and chairs.
GroWrk
Best for: coverage in Latin America and parts of Asia where the larger providers subcontract.
Why companies choose it: in-country presence, which shows up directly in delivery times and in whether a retrieval happens at all.
Where it struggles: publishes nothing, and names its a la carte and Flex models without attaching figures.
allwhere
Best for: United States-led companies with international staff and real configuration requirements.
Why companies choose it: depth in deployment rather than only logistics, which matters when machines need to arrive set up.
Where it struggles: consultation-only pricing, so nothing can be compared before a call.
Manufacturer programmes
Best for: large standardised fleets in major markets.
Why companies choose it: a single supplier for hardware, warranty and service, with support infrastructure that no software-led provider can match at scale.
Where it struggles: one catalogue, thinner coverage in smaller markets, and contract structures built for enterprise procurement rather than for a 200-person company. Pricing is not published and, in Lenovo's case, the TruScale URL carried in older comparisons now 404s, so verify where the current programme sits before relying on any figure attached to it.
What Each One Published
| Option | Published price | Unit | What you are mainly buying |
|---|---|---|---|
| RemoAsset | Not published, demo required | n/a | Lifecycle, issue through recovery |
| Workwize | Not published | n/a | Regional storage and reissue |
| Deel IT | Not published | n/a | Equipment inside employment |
| Firstbase | Not published | n/a | Whole home setup |
| GroWrk | Not published, models named only | n/a | Emerging-market coverage |
| allwhere | Not published, consultation | n/a | Deployment depth |
| Manufacturer programmes | Not published | n/a | Scale, warranty, single supplier |
Checked against each vendor's own page on 6 and 7 October 2026.
The Decision Table
| Situation | Scale | Setup | Primary Pain | Recommended Starting Point |
|---|---|---|---|---|
| Buying 20 machines every 2 years | Under 50 | One office | No real failure yet | Purchase outright, keep spares |
| New hires waiting weeks for a laptop | 50 to 300 | Remote | Deployment, not hardware | A remote lifecycle platform |
| Hiring where you have no entity | Any | Multi-country | No route to buy locally | A platform with direct coverage there |
| Capital spend is the blocker | Any | Any | Balance sheet, not operations | Financing with services, named as such |
| Devices never come back | 50 plus | Remote | End of life, invisible until counted | A platform that takes them back |
| Several thousand devices, few countries | 1,000 plus | Mixed | Unit cost at scale | A manufacturer programme or direct lease |
| Latin America or Asia is the gap | Any | Remote | Coverage, not process | GroWrk or a regional specialist |
| Already running Deel | Any | Remote | One vendor relationship too many | Deel IT inside the existing contract |
Most companies land in two rows, usually a deployment row and an end-of-life row. Those two together are the real case for a subscription, and the hardware is the part you concede on.
What Getting This Wrong Costs
The expensive mistake is not choosing the wrong vendor. It is signing a subscription to solve a hardware cost problem when the actual problem was deployment, and then measuring the arrangement on the wrong thing. Two years later the monthly figure looks high against the price of a laptop, which it always will, and nobody has counted the weeks of coordination or the devices that never came back. The arrangement gets cancelled, the old problems return, and the lesson drawn is that subscriptions are poor value.
The second cost is end-of-term surprise, and it is entirely avoidable. Return conditions, refresh triggers and buyout figures are all knowable on day one and are almost never asked for, because the evaluation is conducted by people solving a day-one problem. A company that asks what month 36 looks like, in writing, before signing, is a company that will not be surprised.
The third is lock-in that nobody priced. Once a provider holds your device records, your regional stock and your retrieval workflow, moving is a project rather than a decision. That is not an argument against the model, it is an argument for asking during the evaluation what leaving looks like, how your asset data comes out, and what happens to devices in their warehouses. The answers are usually reasonable, and asking is what keeps them that way.
So ask the diagnostic question plainly. Are you buying hardware, buying deployment, or buying an exit from capital expenditure? Each of those points at a different shape, and the one thing all three have in common is that the monthly figure is the least informative number in the proposal.
When You're Ready to Move Beyond Buying Outright
The signals are specific. New starters are waiting more than a week for a working machine, and it has happened more than once. You are hiring into countries where you have no route to buy locally. You counted what failed to come back last year and the number was uncomfortable. And somebody will own the vendor relationship through the renewal rather than only the purchase.
When those are true, a subscription usually pays back through deployment and recovery rather than through hardware, and framing it that way internally prevents the wrong comparison later.
If the loop you are trying to close runs from purchase through to return, RemoAsset is built around that specific shape, which is why it suits teams whose devices leak at the end and why it is the wrong tool if what you actually want is a financing instrument. It is worth a look alongside the alternatives here, and expect a demo rather than a price list from every one of them.
And before any of it, ask the vendor what month 37 looks like. The answer tells you more about the agreement than anything on the first page, because it is the only part of the proposal written for their benefit rather than yours. A provider who answers it plainly, with figures and an inspection standard, is a provider who expects you to still be there. One who treats it as a detail to settle later has told you something useful at no cost.
Frequently Asked Questions
What is device as a service?
Device as a service is an arrangement where a provider supplies hardware to your staff on a subscription rather than selling it to you, and bundles services around it. Those services typically cover deployment to the individual, support and replacement during the term, and collection with data destruction at the end. The monthly figure covers all of it, which is both the appeal and the difficulty, because the components are rarely broken out and they are worth very different amounts to different companies. The useful way to evaluate it is to separate the four parts and ask which of them you would otherwise do yourself, and how well.
Is device as a service cheaper than buying laptops?
On hardware alone, no, and any provider suggesting otherwise is describing financing rather than savings. A subscription spreads the cost and adds a margin, which is how the model works. The comparison changes once you cost the parts that do not appear on an invoice: the time spent arranging international delivery, the productivity lost when a machine fails somewhere with no local support, the devices that never come back, and the end-of-life handling you may currently not be doing at all. Companies with a concentrated domestic fleet usually find buying is cheaper. Distributed teams usually find the opposite, and the difference sits almost entirely in those unpriced items.
What happens at the end of a device subscription?
Three outcomes are normal: return the device, extend the term on it, or buy it out at a stated figure. The terms for each are set in the agreement and are the single most overlooked part of the evaluation, because the decision is made by people solving an immediate problem. The specific things to establish before signing are the return condition standard and who applies it, whether the term runs from contract start or from each device's deployment date, and what a buyout costs at both the midpoint and the end. All three are knowable on day one and all three determine whether the arrangement was good value.
How is device as a service different from leasing?
Leasing is a financing instrument covering the hardware, while device as a service bundles financing with deployment, support and end-of-life handling. In practice the line blurs, because lessors add services and service providers add financing, and the label a vendor uses tells you less than the contents. The practical distinction is where the value sits: if you would do your own deployment and recovery perfectly well and simply want to spread the cost, a lease is the more honest and usually cheaper instrument. If the operational work is the thing you cannot do, the bundle is what you are buying and the financing is incidental.
How much does device as a service cost?
No provider in this comparison publishes a figure, which was checked against each one's own pricing page on 6 and 7 October 2026. RemoAsset, Workwize, Deel IT, Firstbase, GroWrk and allwhere all route to a demo or consultation, and the manufacturer programmes quote through enterprise procurement. Two vendors describe a model without numbers: GroWrk names a per-order option alongside a subscription tier, and allwhere refers to pay-as-you-go and fixed rates. Expect pricing to vary considerably by country and by volume, which is the usual reason a category withholds figures, and build a sales conversation into your timeline accordingly.
Does device as a service include MDM?
Usually not, and assuming it does is a common and costly misreading. These providers put a configured machine into somebody's hands and take it back afterwards; enforcing policy on the device while they hold it, pushing settings, requiring encryption and managing applications, is mobile device management and is a separate product. Some providers will enroll a device into your MDM during deployment, which is useful and is not the same as being one. If device policy enforcement is a requirement, confirm specifically what the provider does and does not do before relying on it, and budget for the MDM separately.
What happens if headcount falls during the term?
That depends on the agreement, and it is the clause to read most carefully because almost every proposal is written around growth. Minimum commitments are normal and not unreasonable, but they mean a reduction in headcount does not automatically reduce the bill, and companies discover this in exactly the quarter they can least afford to. Ask for the specific terms covering a meaningful reduction, read them assuming a bad year rather than a good one, and make the answer a factor in choosing between providers rather than a detail settled after selection.
Who is responsible for data destruction at the end?
It should be the provider, and the agreement should say so explicitly along with what evidence you receive. The distinction that matters is between a device being wiped and your holding documentation that it was wiped to a recognised standard, because only the second is useful when somebody asks. If your sector involves audits or customer security reviews, make documented destruction a shortlisting requirement rather than a question asked after selection, and confirm whether certification is included in the monthly figure or charged separately, since practice varies between providers.
HROpsLab takes no vendor money and publishes no paid placements, which is why this page says "not published" seven times rather than estimating.