Spare Parts Inventory Management: What to Hold, What to Skip

Spare Parts Inventory Management: What to Hold, What to Skip

Walk into most maintenance stores and you'll find both failure modes at once. On one shelf, £40,000 of parts nobody has touched in three years, bought for machines that left the building in 2021. On another, an empty hook where the one seal that stops the packing line should be. The team is over-stocked and under-stocked at the same time, and both problems have the same root: nobody decided what the store was for.

Spare parts inventory management is the set of decisions that fixes that. Not the spreadsheet, not the barcode scanner, not the reorder formula. Those are mechanics, and the parts inventory template covers them. This post is about the layer above: which parts earn a place on the shelf, how many, who controls them, and how you know whether the whole thing is working.

What Spare Parts Inventory Management Actually Decides

Most guides on this subject jump straight to reorder points and ABC analysis. Useful tools, but they answer the second question before the first. There are three decisions, in order:

1. Hold it or don't. For every part, is there a reason to tie up cash and shelf space, or can you buy it when you need it?
2. How many, and when to reorder. Only relevant for the parts that passed question one.
3. Who controls it. Where it lives, who can issue it, and what they must record when they do.

Get question one wrong and the other two just optimise a bad list.

Start With the Asset, Not the Part

A parts list built by walking the store and writing down what's there is a list of past purchasing decisions. A parts list built from your assets is a plan.

Take your critical assets first. If you haven't ranked them, the riskiest assets in your business walks through how. For each one, write down the components that fail: belts, seals, bearings, filters, contactors, sensors, the gearbox, the drive. That's the asset's bill of materials, and it's the only honest starting point for a spares list, because it tells you what each stockout actually threatens.

This is why the "used on" relationship matters more than any other field in your parts register. A part that shows Used on: CONV-02, CONV-03 carries its own justification. A part with that field blank is a part you can't defend holding. It's the same discipline as asset definitions in a CMMS: the structure you build at the start decides what questions you can answer later.

Twenty critical assets with proper BOMs beat three hundred with nothing. Work down the criticality list as time allows.

The Stock-or-Skip Matrix

With the BOMs in hand, each part gets one decision. Two factors drive it, and neither is price.

The first is supplier lead time: how long between noticing you need it and having it in your hand. Same day from the local factor, or eight weeks from a manufacturer in Germany.

The second is failure consequence: what happens while you wait. A noisy fan in an office, or a stopped production line, or a safety system out of service.

Put them together and the policy falls out:

The bottom-right cell is where the money and the arguments are, so it gets its own section below.

A worked example, with illustrative numbers. A conveyor gearbox costs £3,200 and has a six-week lead time. The conveyor feeds the only packing line, and a lost shift costs roughly £9,000 in missed output. The gearbox has failed once in seven years. On a usage-based reorder calculation it never qualifies for stock. On the matrix it's an insurance spare without debate: one failure pays for the part nearly three times over in the first shift. The decision isn't about how often it fails. It's about what it costs you when it does.

Insurance Spares Are Not Slow Movers

This distinction is the most common mistake in spare parts inventory management, and it costs teams twice.

A slow mover is a part you stocked expecting to use and didn't. It's a purchasing error, and the right response is to stop reordering it and let it run down.

An insurance spare is a part you stocked hoping never to use. It sits there for years by design. On a standard "zero issues in 24 months" report it looks identical to the slow mover, and that's how it ends up in the skip during a storeroom clear-out, six months before the gearbox it was bought for fails.

So flag insurance spares explicitly in your register. Give them a category, exclude them from dead-stock reports, and give them their own maintenance. Stored motors need their shafts turned; seals and belts have a shelf life; boards want dry, anti-static storage. An insurance spare that's spent five years in a damp corner is a £3,200 placebo. Put an annual inspection on each one, exactly as you would for the asset it protects.

Make Every Issue Cost Something

Parts leave the store against a work order, or they don't leave. That one rule does more for your inventory than any software feature, and it only works if it's absolute. No "I'll book it out later," no "it's only a filter."

When every issue carries a work order reference, three things become possible:

• Cost per asset. After six months you can see that PUMP-07 has eaten £600 in seals, which is evidence a repair-or-replace decision can actually use.
• Real consumption rates. Your min levels stop being guesses and start reflecting what the plant uses, by asset, by season.
• Leakage shows up. When the stocktake finds four bearings missing with no work order to account for them, you know there's a path out of the store that bypasses the log.

The mechanics are in the work order management guide. The short version: the parts line is part of the work order, not a separate form, and the technician records it at the point of use, not from memory on Friday.

Obsolescence and the Annual Cull

Dead stock accumulates in the gap between an asset being decommissioned and anyone telling the storeroom. Once a year, run the cull as a structured exercise rather than a tidy-up:

• Pull every part with zero issues in the last 24 months.
• Remove anything flagged as an insurance spare from that list before you look at it.
• For each remaining part, check the "used on" field against your live asset register.
• Linked asset gone? Sell it back, sell it on, or scrap it. Write off the value and move on.
• Linked asset still running but non-critical? Let the stock run to zero, then buy on demand.
• Linked asset still running, critical, with a long lead time? You've just found an insurance spare that was never classified. Flag it.
• "Used on" blank? That's a register gap, not a stock decision. Fix the record first.

The cull's real output isn't the skip full of parts. It's the list of register corrections, because each one is a stocking decision that was being made by default rather than on purpose.

Measuring It: Five Numbers, Read by Tier

The temptation is to track inventory turns for the whole store and call it a KPI. Turns are a fair measure for the "stock to usage" cell of the matrix and misleading everywhere else. A critical spare that turns over every few months isn't an inventory success; it's a reliability problem wearing a good disguise.

Measure by criticality tier instead, and keep the list short:

Notice what's missing: total inventory value, and store-wide turns. Both move for reasons that have nothing to do with whether the right part is there when a machine stops. These five sit inside a broader set of maintenance KPIs; they're the ones that tell you whether the store is doing its job.

Who Owns the Store

One person owns the register. Not the department, not "whoever's in": one named person who approves additions, runs the annual cull, and signs off the stocktake. Without that, the parts list becomes a shared drive. Everyone contributes, nobody curates.

For field teams, treat each van as a small store that moves. Define its kit, replenish it against work orders, and list it in the register as a location. Otherwise six vans end up holding eight of a part the company needs two of, and still run out.

When to Move It Into Your CMMS

A well-kept spreadsheet handles one store and one keeper. The cracks appear at predictable points: more than one person issuing parts, stock across sites or vans, a need for same-day reorder triggers rather than a weekly review, or anyone asking for cost per asset without a monthly pivot-table ritual.

At that point the parts register belongs inside the same system as the work orders, so the issue is recorded from the technician's phone as the part goes on the machine, and the count, the cost, and the asset history update from that one action. That's how work orders in AssetOS handle it. The decisions in this post don't change when you move; they just stop depending on one person remembering to update a file.

Shane Price

AssetOS

Writing about maintenance management, CMMS implementation, and the real challenges operations teams face.