The switching point between buying and subscribing is not a headcount. It is three variables: how often you refresh, how much churn you have, and whether anybody is currently doing the recovery and reissue work at all.
TL;DR
- Fleet size barely matters. A 60-device fleet with high churn across many markets is a stronger case than 400 devices in one office.
- A short refresh cycle favours subscribing, because you are paying for the part of the life you actually use.
- High churn favours subscribing only if the contract stops billing on early return. Check before assuming.
- If nobody does the recovery work today, buying is cheaper on paper and more expensive in practice.
- Owning and holding spares in region beats both arrangements for day-one readiness in slow markets.
- Nobody publishes a price, so you cannot model this without running a sales process first.
Why headcount is the wrong variable
Device count determines the size of the number, not which arrangement wins. A larger fleet makes both options bigger in proportion and the comparison between them barely moves.
What changes the answer is how much of each device’s life you consume and how much of the lifecycle work you are actually doing. A company that buys machines, uses them for 5 years and disposes of them properly has a cheap arrangement. One that buys machines, loses track of a third of them and refreshes irregularly has an expensive one that looks cheap in the budget.
So the comparison is not purchase price against subscription. It is total lifecycle cost including the work, against a subscription that includes some of that work.
The three variables
| Refresh interval | Short favours subscribing | At 3 years you use most of the value. At 5 you have paid for life you did not need |
| Churn | High favours subscribing, conditionally | Only if billing stops on early return. Otherwise it favours nothing |
| Who does the work today | Nobody favours subscribing | Recovery, wiping, storage and reissue are costs whether or not anybody is doing them |
The third variable is the one that decides most real cases and it never appears in a cost model, because unpaid work that is not happening has no price. A company with a 30 per cent recovery rate is absorbing the cost of the other 70 per cent and calling buying cheap.
Where buying stays better
Long refresh intervals with a stable team. If machines last 5 years and people stay, you are capturing nearly all of the asset’s value and a subscription is paying somebody else for a service you do not need.
One location with a store cupboard. Recovery is a walk down a corridor, storage is a shelf, and reissue is handing somebody a machine. The work a subscription outsources costs you almost nothing.
A functioning spare pool. If recovered machines are triaged, wiped and reissued, you already have the main benefit of a lifecycle service without the fee.
Capital availability and a preference for owning. A legitimate position, and worth stating plainly rather than dressing up as an arithmetic conclusion.
A worked example
Two companies compared the same decision. Both bought machines and both were considering a subscription.
The first had 410 devices in two offices, a 4-year refresh, low turnover and a store cupboard with a named owner. Recovered machines were wiped on arrival and reissued. For them buying was clearly cheaper and the subscription offered nothing they were not already doing.
The second had 62 devices across 9 markets, no office, 3-year refresh and 40 per cent annual turnover. It recovered about a third of departing devices, held no spares, and bought new for every starter because nothing reusable was ever in the right country. Buying looked cheap in the budget and was substantially more expensive once the unrecovered machines and the unnecessary purchases were counted.
The smaller fleet was the stronger case for subscribing, which is the opposite of what a headcount threshold would suggest.
Working out your own answer
What to do:
- Count your actual recovery rate over the last two years, not your intended one.
- Count machines bought for starters in markets where a reusable device already existed somewhere.
- Write down your real refresh interval, measured from the register rather than from policy.
- Count the markets you hire into and how many have no local storage option.
- Add the unrecovered devices and the unnecessary purchases to the cost of buying, then compare.
- Then get quotes, because no provider publishes a price and the comparison cannot be modelled without them.
The second item is the one that produces the surprise. A company buying new machines in markets where it already owns idle hardware is paying twice, and that figure belongs in the buying column rather than being treated as an operational irritation.
Workwize, Deel IT, Firstbase, GroWrk and allwhere all operate here and none publishes a price. Disclosure: RemoAsset is owned by the same people who publish PeopleOpsHQ. It also publishes no price and requires a demo, it is much weaker for hardware it did not supply, and it is not a certified disposal vendor. The options are compared in our device-as-a-service comparison.
Final thoughts
There is no device count at which subscribing starts to win. The variables are your refresh interval, your churn and whether the lifecycle work is being done, and a small distributed fleet with high turnover is a far stronger case than a large office-based one.
Count your real recovery rate and the machines you bought unnecessarily, put both in the cost of buying, and then compare. If buying still wins, the cheapest improvement available is a regional spare pool funded from recovered machines, which captures most of a subscription’s operational benefit at no fee.
Frequently asked questions
At what fleet size does device-as-a-service beat buying?
There is no such size, and headcount is the wrong variable because it scales both options in proportion. The three that decide it are the refresh interval, since a short cycle means you consume most of a device’s value and a long one means you paid for life you did not need; the churn rate, which favours subscribing only if billing stops on early return; and whether anybody is currently doing the recovery, wiping, storage and reissue work, because those are costs whether or not they happen.
Why can a smaller fleet be a stronger case for subscribing?
Because the case rests on how much lifecycle work is going undone rather than on volume. A sixty-device fleet across nine markets with no office, high turnover and a one-third recovery rate is absorbing the cost of every unrecovered machine and buying new for starters in markets where it already owns idle hardware. A four-hundred-device fleet in two offices with a store cupboard and a named owner is already capturing that value, so a subscription adds a fee for work it is doing well.
When is buying laptops clearly better?
Four situations. A long refresh interval with a stable team, where you capture nearly all of the asset’s value. One location with a store cupboard, where recovery is a walk down a corridor and the outsourced work costs you almost nothing. An existing spare pool where recovered machines are triaged, wiped and reissued, since you already have the main operational benefit. And a straightforward preference for owning with the capital available, which is a legitimate position worth stating as one.
Does high turnover favour a subscription?
Only conditionally, and the condition is in the contract. A subscription priced per device per month assumes a stable count, so if billing continues on a machine returned in month 8 of a 36-month term, churn costs you under either arrangement. Ask specifically whether billing stops on early return before treating turnover as an argument for subscribing, because the clause is frequently absent entirely rather than unfavourable, which means it gets accepted by default.
What costs get left out of the buying column?
Two, and both are large. The devices that were never recovered, which is the difference between your intended recovery rate and the actual one over the last two years. And the machines bought new for starters in markets where a reusable device already existed somewhere in the company, which is paying twice and usually gets logged as an operational irritation rather than a cost. Add both to the cost of buying before comparing anything.
What is the cheapest alternative to a subscription?
A regional spare pool funded from recovered machines, held in the markets where you hire at least three or four times a year. It captures most of a lifecycle service’s operational benefit, which is having a usable machine in the right country, at no fee, because the devices are ones you already own and were holding anyway. Wipe them on arrival, test one against the real applications, label the result, and put a review date on each so the pool does not become a cupboard.