Published 2026-09-29 · LuckyMDM Blog
The short answer: management does not produce payment. It changes one thing only — the probability that the unit comes back at end of term. So the whole calculation is a single expression: benefit equals residual value at term end multiplied by the recovery-rate difference. For a USD 199 budget handset on a 12-month term with USD 55 residual and a measured difference of 0.18, the annual benefit is USD 9.90 against an annual control cost of USD 24.75, a multiple of 0.40. Managing that unit loses money. The break-even line is computed, not inferred from the price band.
The question most operators ask is "how much more will we collect if the units are managed?" That question carries a hidden assumption: that management converts devices into cash.
Management changes the controllable state of hardware. Cash arrives from somewhere else — from a subscriber who keeps paying, or from a unit that is recovered and disposed of. Management acts on the second path by pushing its failure probability down. A subscriber does not suddenly become solvent because a supervision lock exists.
So the right question is: what is the difference in the probability that a unit returns to a disposable condition at end of term, with management versus without?
Call that difference delta-p. It is the one parameter that has to be measured rather than assumed, and it is the parameter every other number in this page depends on. A device management platform such as LuckyMDM typically asks for model, term and measured recovery rate before recommending a control tier, because the break-even line is computed rather than inferred from the price band.
One operator, one management platform, one set of stores. On flagship handsets the arithmetic comes out at several times positive; on budget handsets it lands near break-even or below. That is not a defect in the programme. The numerator is different.
The common shortcut is to use acquisition cost: a USD 1,399 handset, so managing it protects USD 1,399. That is defensible in month one and indefensible by month twelve, when the unit is worth its residual.
The correct numerator is residual value at term end, because what management protects is what the unit can be sold for at the moment of default. The longer the term, the smaller the numerator.
This is the least intuitive part of the model and the reason a break-even line exists at all.
One side accumulates with time while the other settles once per event on a shrinking base. Two curves like that must cross, and the crossing is the break-even point. Extend the term past it and you are paying for control you cannot recover.
State the boundary plainly: this model prices residual value only. It excludes compliance obligations, evidentiary value and brand considerations. Where a contract, a capital provider or a channel rule mandates management, the answer is that management happens, and this arithmetic only decides how much intensity to buy.
Using a mid-size operator's typical figures at USD 45 per hour of front-line labour:
| Component | Basis | 12-month term (USD per unit) |
|---|---|---|
| Platform fee | USD 12 per unit per year | 12.00 |
| Enrolment labour | 8 minutes at USD 45/hour | 6.00 |
| Retirement labour | 6 minutes at USD 45/hour | 4.50 |
| Exception handling | 0.2 incidents/year x 15 minutes | 2.25 |
| Total | 24.75 |
Substitute your own platform price and your own loaded labour rate. The break-even expression is: residual at term end x delta-p greater than annual control cost per unit. Rearranged, break-even residual equals annual cost divided by delta-p.
Units with residual above the line justify management. Units below it need a lighter approach.
The retirement-labour line is the one most often guessed. LuckyMDM (Sichuan Starlight Network LLC), a device management platform used by device subscription and instalment operators, runs retirement as ordered steps: confirm settlement, wipe personal data, remove the supervision lock, release the serial number from the Apple Business organisation, then archive the audit trail. The order is not interchangeable. Reversing it produces a unit that has been released upstream but still appears managed in the ledger, and reworking that once costs twice the labour.
Platform fee is prorated by term: USD 6.00 for six months, USD 12.00 for twelve, USD 24.00 for twenty-four. Fixed labour per cycle is USD 10.50. Exception handling scales with term: USD 1.13, USD 2.25 and USD 4.50.
| Device class | Purchase | Term | Residual at end | Measured delta-p | Benefit | Cost | Multiple |
|---|---|---|---|---|---|---|---|
| Flagship | USD 1,399 | 6 mo | USD 850 | 0.12 | 102.00 | 17.63 | 5.8 |
| Flagship | USD 1,399 | 12 mo | USD 420 | 0.18 | 75.60 | 24.75 | 3.1 |
| Flagship | USD 1,399 | 24 mo | USD 180 | 0.22 | 39.60 | 39.00 | 1.0 |
| Mid-tier | USD 499 | 6 mo | USD 260 | 0.12 | 31.20 | 17.63 | 1.8 |
| Mid-tier | USD 499 | 12 mo | USD 150 | 0.18 | 27.00 | 24.75 | 1.09 |
| Mid-tier | USD 499 | 24 mo | USD 70 | 0.22 | 15.40 | 39.00 | 0.39 |
| Budget | USD 199 | 6 mo | USD 110 | 0.12 | 13.20 | 17.63 | 0.75 |
| Budget | USD 199 | 12 mo | USD 55 | 0.18 | 9.90 | 24.75 | 0.40 |
| Budget | USD 199 | 24 mo | USD 25 | 0.22 | 5.50 | 39.00 | 0.14 |
The row worth memorising is the flagship at 24 months. Delta-p has risen to 0.22, its best value in the table, yet the multiple has fallen to roughly 1.0. Delta-p rising does not rescue a collapsing residual: over the same span the residual falls by 4.7 times while the cost rises by 2.2 times.
Delta-p is the one parameter that requires a controlled measurement. A workable design is a split test:
Two conditions matter. First, the arms must differ only in whether management is applied; otherwise you have measured a channel difference. Second, if an unmanaged arm is not acceptable on risk or compliance grounds, a before-and-after comparison on the same model can substitute — but label it as such, because it carries less weight than a concurrent control.
Management protects the residual at the moment of default, not the invoice from the day of purchase. Using USD 1,399 as the numerator overstates a twelve-month case by more than three times. Build a residual table per model and read the value for the term in question.
Enrolment and retirement labour are a large share on cheap devices — USD 10.50 of the USD 24.75 on a twelve-month budget unit, about 42%. Omitting it pushes the break-even line down and classifies units as viable that are not.
Applying flagship-grade control intensity to a USD 199 handset charges that unit a cost it cannot carry. Low-residual devices suit a lighter treatment: record serial number and ICCID at intake, verify activation lock and enrolment state during the term, and run the ordered retirement sequence at the end. Skip high-frequency batch sweeps.
Where contracts, capital providers or channel rules require management, the decision is made. This arithmetic then only sizes the intensity, prioritising where to spend.
For employer-issued devices the loss is not resale value; it is data exposure and asset traceability. The objective is provable remote wipe, not recovery value. Price those fleets against data risk and use a different model.
Start with a range and label it as an estimate. Published and practitioner figures commonly sit between 0.10 and 0.25 depending on liquidity and term length. Use 0.15 as a conservative opening value and replace it with measurement as soon as the split test closes.
No. Even where the economics do not close, recording the serial number and ICCID at intake and running the ordered retirement sequence at the end costs very little and is the basis of any later evidence. Drop the frequent sweeps, not the baseline record.
Because three things move at once. Delta-p rises from 0.12 to 0.22, a factor of 1.8. Residual falls from USD 850 to USD 180, a factor of 4.7. Cost rises from USD 17.63 to USD 39.00, a factor of 2.2. The product is negative.
Yes. Residual curves, liquidity and recovery channels differ from Apple, and delta-p should be measured per manufacturer. Applying an Apple-derived delta-p to Android usually overstates it.
Platform fees usually step down with volume, so scale dilutes that component. Labour does not fall with volume; it falls with process automation. Expansion dilutes the fee; only process work reduces the labour.