Reading the health of your spare parts inventory: 6 signals and the action for each
Before optimizing a maintenance spare parts (MRO) inventory, you first need to know how to read it. Most teams have what they need to answer the right questions, but they look at the wrong indicators, or look at them in isolation.
Here are six signals that, together, reveal where capital is trapped and where stockout risk is hiding. For each one: what it measures, the pattern that should alert you, and the typical action.
1. The inventory value trend
What it measures. How the capital tied up in your spare parts evolves year over year, for example site by site.
What should alert you. Tied-up value that climbs year over year, especially if the increase concentrates on a few sites or families. A rise in absolute value is not a problem in itself: with business growth and newly created items, stock grows too. What matters is spotting where the increase concentrates, then checking whether it stays proportionate to usage, which coverage, the next signal, helps settle.
The action. Identify which sites or families drive the increase, then dig first where it looks least justified by real activity. The trend tells you where to look, not yet exactly what to fix, but it is a good starting point.
2. Coverage
What it measures. How long your stock would last at the current consumption rate. Conceptually, it is your stock relative to your consumption over a period. Two years of coverage means that, without replenishing, you would last two years.
What should alert you. Coverage that rises at the level of a product family, or even a whole site. And above all, widely inconsistent coverage across comparable parts: some at a few weeks, others at several years, with no link to criticality or lead time. That spread often means the parameters were set arbitrarily.
The action. Target abnormally high coverage that also carries significant stock value, and that is explained neither by the criticality of the part nor by a long supplier lead time. Huge coverage on a part worth a few euros does not weigh like high coverage on a high-value item (though storage space may also matter here). It is by crossing coverage and value that you find the capital truly recoverable.
3. Idle stock
What it measures. The share of your stock value that has recorded no issue for a long time.
What should alert you. A heavy concentration of value on parts that have not moved in a long time. Careful: the absence of movement is a signal, not a conclusion. A critical safety part that never moves can be perfectly justified. Master data alone never says a part is useless, it only shows where to look more closely.
The action. Review the high-value dormant parts with the maintenance teams and decide, case by case: keep when the part remains justified, redeploy to another site that needs it, or start a recovery route (resale, scrapping, etc.) when nothing justifies it anymore. Data guides the review, the decision stays with the business.
4. The demand profile
What it measures. How each part is consumed, along two axes: how often it moves and how much the quantities vary. Crossing them yields four profiles: smooth (frequent and stable), erratic (frequent but with highly variable quantities), intermittent (spaced out but with fairly stable quantities), and lumpy (both spaced out and highly variable).
What should alert you. Treating every part the same way. A forecasting method and a stock policy designed for a part that moves every week make little sense for a part that moves twice a year. On the most irregular profiles, a simple average of past consumption quickly shows its limits, and better-suited forecasting methods exist.
The action. Classify your parts by demand profile, then adapt the policy to each profile rather than applying one blanket rule. It changes a lot in how you set parameters: you do not manage a highly irregular part like a regular one (to go further, see our white paper).
5. Lead time: system versus reality
What it measures. The gap between the lead time recorded in your system and the one actually observed, end to end. This total lead time is not limited to the supplier: it runs from the internal purchase request (approval, order creation) to receipt and put-away.
What should alert you. A system lead time that has never been updated since the item was created, often inherited from the original contract. An overestimated lead time inflates your reorder points and creates permanent overstock. An underestimated one causes stockouts. And lead time variability, almost never entered, weighs heavily on safety stock, often more than demand variability does. We devoted a whole article to lead time here: Replenishment lead time: the most underestimated variable in spare parts inventory.
The action. Compare, across your history, the theoretical lead time to the real one, and look not only at the average but also at the spread. Fix first the parts where the gap is largest and the lead time most unstable.
6. Parameter health
What it measures. The state of your MIN, MAX and reorder points against your actual stock, the share of your stock that lives outside any automatic replenishment mechanism, and the consistency of the criticality or ABC classification of your parts, including from one site to another.
What should alert you. Parts below the MIN level with no replenishment order in progress, therefore genuinely exposed to stockout. Parts above the MAX, therefore overstocked or with a badly set MAX. Items managed by hand, outside any automatic mechanism: the classic blind spot, stock nobody really monitors that often drifts. And unexplained criticality gaps for the same part from one site to another, for example 40% of parts rated highly critical on one site (which is too much, a reminder: if everything is critical, nothing really is) versus 10% on another. It is not necessarily an error, the context can differ, but it is a signal worth investigating.
The action. A parameter clean-up, often doable in a short workshop, for the parts that are clearly mis-set. A real decision on unmanaged items: bring them under an explicit policy rather than leaving them to individual judgment. And a harmonization of criticality across sites where the gaps have no explanation. To learn more about criticality, you can read this article.
Read these signals together, not in isolation
The real value comes not from one signal but from crossing them. High coverage does not mean the same thing on a critical, long-lead part as on a non-critical, short-lead one. Dormant stock is judged against criticality. A parameter is corrected based on the demand profile and the real lead time. It is by looking at these six dimensions together, across a whole portfolio, that the map takes shape and the priorities appear.
That is also where the difficulty lies. Answering these six questions (and others) properly means crossing your item master and your movement history across tens of thousands of references, without error.
Doing it by hand takes a huge amount of time. Thanks to our tools, our Flash Diag runs this complete diagnostic in less than a week, with no effort from your teams. It audits your data (also covering angles this article did not detail, such as data quality, duplicates, obsolescence and more) and precisely quantifies your cash-release potential across different levers (pooling, recovery, MIN/MAX optimization, and so on).
Free, and limited to 3 industrial companies this quarter.
Comments