Slow-moving and obsolete stock: the money already on your shelves

Every maintenance store has a shelf, usually high up or right at the back, holding parts for a machine that was decommissioned in 2019. Everybody knows roughly what it is. Nobody moves it, because moving it means deciding, and deciding means somebody has to be the person who threw away a part that turned out to be needed.

So it stays, and the same reasoning applies to the next machine, and after a decade a meaningful fraction of the store is stock that will never be issued.

What it actually costs

The instinct is that it costs nothing, because the money was spent years ago. That is a sunk cost, and it is exactly the reasoning that keeps it there. What it costs now:

• Space, which is the one people feel. A store that is full needs an extension, or it needs a clear-out.
• Search time. Every item in the store is something to look past, and a store where half the shelf is dead stock is a store where finding things is slower for everyone, every day.
• Counting effort. Cycle counting dead stock is pure waste — you are verifying a number nobody will use.
• The balance sheet, if it is valued. Stock carried at cost that will never be issued overstates current assets, and an auditor will eventually say so.
• Occasionally, actual risk: perished rubber, expired adhesives, degraded seals that are still on the shelf and still look fine.

The question is not "did we waste money buying it". That is settled. The question is what the shelf is worth to you now.

Finding it without a project

You need one number per line: when it was last issued. Not last counted, not last received — last actually taken by somebody for a job. If your system records movements rather than only balances, this is a single report and it takes minutes.

That last row is the important reversal. For genuinely dead stock the burden of proof should sit with keeping, not with disposing — otherwise nothing ever leaves, because "we might need it" is unfalsifiable and always available.

The insurance spares problem

There is one legitimate category that breaks the rule above, and confusing it with dead stock is how organisations end up scrapping something expensive and then buying it again.

An insurance spare is a part deliberately held for a critical machine, with a long lead time, where the cost of the outage vastly exceeds the cost of holding it. A pump impeller for the only line in the plant, with a fourteen-week lead time, is not slow-moving stock. It is a decision that was made once and should be recorded as such.

The fix is straightforward and almost nobody does it: flag them. Mark insurance spares explicitly, with the machine they protect and the reason. Then a slow-moving report can exclude them, and a review does not have to relitigate the same argument every year. If a part is unflagged and has not moved in four years, it is not an insurance spare — it is stock somebody forgot.

Deciding, in order of what to try first

1. Use it. Obvious, and often possible — a bearing bought for a scrapped machine may fit three others. Search by specification, not by the part number it was bought under.
2. Move it. Another site, another division, a sister company. This is where a multi-site view earns its keep, because the part is only dead where it is standing.
3. Return it. Some suppliers take back unused stock, especially if you buy from them regularly. Rarely offered, occasionally accepted, always worth asking for anything substantial.
4. Sell it. Specialist obsolete-parts dealers exist for exactly this, and for older industrial kit the market is better than people expect.
5. Scrap it, and record why. Value recovered from metals is not nothing, and neither is the shelf.

Write-off is a record, not a bin

The part people get wrong: a write-off should leave a trace. What was disposed of, when, on whose authority, and for what reason. Deleting the line from the system is not a write-off — it is a hole where the history was, and it makes the next question ("didn’t we used to hold those?") unanswerable.

Keeping the record also lets you check yourself later. If you scrapped something and bought it again within a year, that is worth knowing about before the next review, and it is the only way the threshold above ever gets calibrated to your actual business rather than to a table in an article.

Then stop making more

The clear-out is the easy half. Dead stock is created continuously, by minimum levels set for consumption that has since stopped and by orders placed for machines that are on their way out. Reviewing reorder points on anything that has slowed down — and clearing them when a machine is decommissioned — is what stops the shelf refilling itself.

Nilstock records every issue as a ledger entry, so "when did this last actually move" is a fact rather than an estimate, and a write-off is a recorded adjustment with a reason rather than a deletion. On the levels that create it in the first place, see setting reorder points; on keeping the balances trustworthy enough to act on, cycle counting. The maintenance stores page covers the spares side specifically.