The replacement price of a lost item is usually the smallest part of its cost. Add the work stopped while it's unavailable, the time spent looking for it, the duplicate purchase it often triggers, and the audit exposure of an inaccurate register. The full figure is commonly two to four times the sticker price.
Ask what a lost thermal camera costs and you'll be told the replacement price. It's the wrong number, and it's wrong by a large multiple.
Here's a method for producing one you can actually defend in front of finance.
The five components
|
Component |
How to estimate it |
Frequently missed? |
|---|---|---|
|
Replacement cost |
Current price of an equivalent unit |
No — it's the only one people count |
|
Search time |
Hours spent looking × loaded hourly cost |
Yes |
|
Work stopped or delayed |
Delay × cost of the delayed activity |
Yes, and often the largest |
|
Duplicate purchase |
Items bought that already existed elsewhere |
Yes |
|
Administrative handling |
Write-off, reorder, approval, insurance |
Yes |
Replacement cost
Use current replacement price, not what you paid. And use book value, not purchase price, when you're arguing about whether to have prevented it — a five-year-old instrument isn't a five-year-old instrument's worth of loss.
Search time
The cost before anyone concludes it's gone. Usually several people, over days, in fragments. The fragments are why it's underestimated — nobody logs twenty minutes.
Measure it rather than guess. Ask a handful of people to log every search for two weeks. Two weeks of your own data is worth more than any benchmark, and it survives scrutiny.
Work stopped or delayed
This one is situational and often dominates everything else.
A missing laptop costs a day of one person's output. A missing calibrated instrument can stop a team, blow a client commitment, or force a hire at short notice — and short-notice hire is priced accordingly. On a construction or field job, the delay cost routinely exceeds the item cost by an order of magnitude.
Duplicate purchase
Two flavours. The replacement for the lost item, and — separately — buying something you already own elsewhere because nobody knew it existed.
The second is easy to find and worth doing before you build the rest of the model: cross-reference last year's equipment purchases against your register and look for items bought in one department that already sat unused in another.
Administrative handling
Write-off approval, reorder, insurance if claimed, updating records. An hour or two of somebody's time per incident, and it scales with your approval process — organisations with heavier governance pay more per loss, which is rarely factored in.
A worked example
|
Component |
Amount |
Basis |
|---|---|---|
|
Replacement |
£1,400 |
Current price of an equivalent unit |
|
Search time |
£180 |
6 hours across 3 people at £30 loaded |
|
Delayed work |
£2,000 |
Half-day team delay plus short-notice hire |
|
Duplicate purchase |
£0 |
None in this case |
|
Admin handling |
£90 |
3 hours across write-off and reorder |
|
Total |
£3,670 |
2.6× the replacement price |
That multiple — somewhere between two and four times sticker price — matches most of the models I've helped build. Your mileage will vary with how much your work depends on the item being available.
Where the multiple comes from
Two to four times replacement cost sounds high until you decompose a real incident. The spread is driven almost entirely by one variable: whether the missing item stopped anyone working.
|
Situation |
Multiple |
Why |
|---|---|---|
|
Spare item, plenty available |
1.2× |
Replacement plus a little admin |
|
Office equipment, one person affected |
1.5–2× |
A day of output plus search time |
|
Field equipment, team waiting |
3–6× |
Delay and short-notice hire dominate |
|
Calibrated or certified item |
4–8× |
Work may need redoing, or can't proceed at all |
Which tells you where to concentrate. Losing a spare monitor is an irritation. Losing the one calibrated tester is a day's team cost plus whatever the client thinks about it, and those items deserve disproportionate attention.
The cost that compounds: a register nobody trusts
There's a threshold — in my experience somewhere around 15–20% inaccuracy — past which people stop consulting the register and start asking around instead.
After that it degrades quickly, because the only thing keeping a register current is people using it. An unconsulted register isn't updated, so it gets worse, so fewer people consult it. You end up with a system you're paying for and a process running on memory.
It's why I'd rather see a small accurate register than a comprehensive stale one, and why measuring accuracy matters more than measuring coverage.
Tracking losses so the number improves
Most organisations record a loss as a write-off line and nothing else, which makes the total visible and the pattern invisible.
Four fields turn write-offs into something you can act on:
- Last seen — date and location. Without this you cannot spot a site or a stage that loses more than its share.
- Last holder — not to apportion blame, but because a role losing repeatedly indicates a process gap in that role.
- Circumstance — one of a short list: not returned, left on site, unknown, damaged beyond repair, stolen.
- Search outcome — found later, or not. A high found-later rate means your register is wrong rather than your equipment gone.
That last one surprises people. On a first pass, a meaningful share of 'losses' turn out to be items that were somewhere the register did not know about — which is a cheaper problem to fix than theft.
The annual view worth building
|
Metric |
What a bad number looks like |
What it points at |
|---|---|---|
|
Loss rate by site |
One site well above the others |
A local process gap, not a people problem |
|
Loss rate by category |
Concentrated in low-value portables |
Registration gaps — the excluded tail |
|
Found-later rate |
Above 30% |
Register accuracy, not loss |
|
Losses at departure |
A visible spike |
Offboarding has no list |
|
Average days to notice |
Measured in months |
Counting frequency is too low |
Any one of those pointing somewhere specific is worth more than the headline total, because the total tells you there is a problem and these tell you where it is.
Reducing the number
- Attack delay first, since it's the largest component. Knowing what's available before you need it prevents the scramble entirely.
- Then search time. Scannable labels and a location that's actually current turn a twenty-minute hunt into a lookup.
- Then duplicates, by making the register visible across departments rather than per-team.
- Replacement cost last. It's the number everyone starts with and the one you can do least about.
The one you can't put a number on
A register with known gaps is a register nobody trusts, and an untrusted register stops being consulted. At that point you're running on memory and asking around, which is the state you were trying to leave.
Don't try to price it. Name it as a risk, note who has asked you for a verified count before, and let it sit alongside the quantified figures.
Key takeaways
- Replacement price is typically well under half the true cost of a lost item.
- Delayed or stopped work frequently dominates, especially in field and construction settings.
- Measure search time for two weeks rather than quoting a benchmark — your own data survives scrutiny.
- Check for duplicate purchases against the register; it's quick and usually finds something.
- Expect a full cost somewhere between two and four times the replacement price.
Frequently asked questions
How do you calculate the cost of lost equipment?
Add replacement cost, time spent searching, the value of work stopped or delayed, any duplicate purchase, and administrative handling. The total commonly lands at two to four times the replacement price, with delay cost the largest variable.
What is the true cost of losing a tool or device?
Well above its replacement price. A £1,400 instrument that delays a team for half a day and takes three people six hours to look for can easily cost £3,500 or more once the delay and the search are counted.
How much equipment do companies lose each year?
Published figures vary so widely they're not worth quoting, and most organisations can't measure it because the items most often lost were never registered. Run a scoped physical count instead — the discrepancy rate is your own number and it's defensible.
Does insurance cover lost equipment?
Often partially, subject to excess, depreciation and whether you can evidence reasonable control. That last point matters: an insurer may ask who held the item and what records you keep, which is where a documented chain of custody earns its place.
How do you reduce equipment loss cost?
Attack the largest components. Reduce search time with scannable labels and accurate locations, reduce delay by knowing what's available before you need it, and eliminate duplicate purchasing by making the register visible across departments.
How do you account for lost equipment?
Write it off at current book value rather than purchase price, record the last seen date and location, and close the custody record rather than leaving it open. The write-off is the accounting entry; the last-seen detail is what lets you spot a pattern across several losses.
Should we charge employees for lost equipment?
This is restricted or prohibited in many jurisdictions without explicit prior consent, and wage deduction especially so. Take advice for your own jurisdiction rather than copying a policy clause. A signed custody record establishes responsibility; whether and how you recover cost is a separate legal question.
Is it worth insuring portable equipment?
Usually, above a certain value — but read what the policy expects of you. Most require evidence of reasonable control, which in practice means being able to identify the item and show who held it. A register answers the first; only a custody trail answers the second.