The claim, and how to check it: Somewhere in your network this quarter, a planner raised an emergency purchase order at a 50-80% premium over standard cost (MRO360 product documentation) for a part a sister plant was already holding in surplus.
That is not bad luck. It is the predictable output of an MRO supply chain fragmented in five specific places: parts invisible across plants, every part treated with equal urgency, dead stock locking capital, reorder points frozen at setup, and procurement running plant by plant.
Each failure has a test you can run this week on one data extract, and a decision it forces. Run the five tests below. If two or more come back hot, you have a quantified case, and closing it takes an overlay on your existing ERP/CMMS, not a replacement, live in 8-12 weeks.
Picture the transaction nobody audits. A line goes down at Plant A. The planner searches local stock: nothing.
Under pressure, they raise an emergency purchase order, expedited freight, premium pricing, a vendor doing them a favor. The part lands in 36 hours and the line restarts.
Everyone did their job correctly. Except the identical part was sitting, unused, in Plant B's storeroom two hundred miles away, recorded under a slightly different description, invisible at the exact moment it mattered.
The enterprise just paid a premium to buy something it already owned. Multiply that transaction across a year and a multi-site network, and you have the real cost of a fragmented MRO supply chain: not one dramatic failure, but a quiet tax collected at five predictable, structural points.
This article names each point, gives you the test that exposes it in your own data this week, and the decision each test forces.
What is MRO in the supply chain?
Thirty seconds of definition, because the trap hides inside it. MRO, maintenance, repair and operations, covers the spares, consumables, and supplies that keep physical assets running, as distinct from the direct materials that go into a finished product.
The MRO supply chain runs from supplier and procurement, through the storeroom and inventory ledger, to the maintenance work order that finally consumes the part. It sits inside industrial supply chain management, but it obeys different physics, and that difference is the reason it needs its own management model.
MRO demand is irregular by nature. Direct materials follow a production schedule you can forecast within a reasonable band: you know roughly how much raw steel next month's output requires, because the demand signal is the production plan itself.
MRO demand is driven by asset condition and failure, not by a plan. A pump bearing might be consumed three times in two years, then twice in one month, because failure clusters around wear, vibration, and operating stress, not around a calendar.
Forecasting tools and reorder rules built for smooth, schedule-driven demand systematically mishandle this pattern; they either average the spikes away and understock, or they pad every SKU with safety stock and overstock. That single distinction, forecastable demand versus failure-driven demand, is why the five root causes below exist.
Why MRO supply chain management fails: five structural root causes
Across the multi-site industrial operators Verdantis works with, in oil and gas, mining, and heavy manufacturing, the same five disconnection points appear again and again, not as occasional glitches, but as structural features of how the estate was originally set up.
Each cause is individually rational at the plant level. The cost only becomes visible at the enterprise level, which is exactly why no single plant ever sees the full bill.
Root cause 1: parts are invisible across plants
A stockout at one site triggers emergency procurement even when the identical part sits in a sister plant's storeroom, because nothing in most ERP and CMMS landscapes makes that surplus visible at the point of need.
Inconsistent naming compounds the problem. The same gasket, recorded under three different descriptions at three plants, is technically "in the system" but practically unfindable by exact-text search.
In one diagnostic on a multi-site manufacturer's own extract, Verdantis found a single material transacted by five separate business units, with three different approved manufacturers listed for what was functionally one interchangeable part. Emergency procurement carries a 50-80% cost premium (per MRO360 product documentation), a premium paid, in cases like this one, for a part the enterprise already owned.
Spot it in your data
Test: Pull last quarter's emergency and expedited purchase orders. Match each part number, and its description variants, against stock on hand at every other site on the same date.
Decision: Every match is a premium paid for a part you already owned. If matches exceed a handful, cross-plant visibility is your first fix, before any stocking policy is touched, because you cannot optimize what planners cannot find.
Root cause 2: every part is treated with the same urgency
Without a structured, part-level criticality model, stocking and prioritization default to gut feel or blanket over-stocking. Most legacy frameworks, FMECA, VED, and ABC classification, model criticality at the asset level, then let every part attached to that asset inherit the asset's rating.
The logic sounds sensible and is backwards in two directions at once. A low-risk, commodity part on a critical asset gets over-protected. A high-risk, single-source, long-lead part on a "non-critical" asset gets ignored, right up until it stops something that mattered.
Part-level criticality has to be computed from part-level facts: supplier lead time, sourcing depth, substitute availability, equipment linkage and redundancy, consumption pattern, and failure consequence, across tens of thousands of SKUs. How that score actually gets built is worth a look on its own.
Spot it in your data
Test: List your top 100 parts by stock value and your top 100 parts by failure consequence (long lead time, single source, installed on critical equipment). Measure the overlap between the two lists.
Decision: If the lists barely intersect, your working capital is protecting the wrong risks. Score part-level criticality before cutting or adding stock anywhere, or every other fix will optimize the wrong parts.
Root cause 3: dead and obsolete stock locks working capital
Discontinued, superseded, or no-longer-linked-to-any-asset parts sit in the storeroom consuming space and capital, usually undetected until a manual audit surfaces them years later. Industry benchmarks cited in MRO360 product documentation put 25-40% of MRO inventory at asset-heavy industrial sites as excess, obsolete, or duplicated.
Real extracts routinely confirm the range. In one Verdantis diagnostic of a multi-mill manufacturer, 42% of inventory value had not moved in 24 months.
The obsolescence question is genuinely hard for internal systems to answer alone, because the decisive facts, whether the OEM still manufactures the part, what supersedes it, live outside the ERP entirely. More on that here, and it is where first deployments typically release 20-35% of working capital, a Verdantis deployment benchmark rather than an industry-wide claim.
Spot it in your data
Test: Compute months-of-cover (on-hand quantity divided by trailing twelve-month consumption) per SKU and plot the distribution. Separately, list every SKU with zero movement in 24-plus months and no active equipment linkage.
Decision: Value above 24 months of cover is your disposition queue: redeploy, sell, or write off with an audit trail. Where cover exceeds 12 months, pause buying immediately.
Root cause 4: reorder points are static
Most reorder points are set once, at plant commissioning or material creation, based on historical average usage, and rarely revisited, even as production volumes, demand patterns, and supplier lead times shift around them for years afterward. The formula makes the problem concrete.
The reorder point formula
Reorder Point = (Average Daily Usage × Lead Time) + Minimum Safety Stock
Average Daily Usageshifts with production and asset age Prazo de entrega
moves with supplier reality Stock de segurança
should follow criticality and variability
All three variables change constantly in reality. In a static model, all three are frozen at whatever they measured on setup day.
The failure pattern this produces deserves to be named precisely: overstock and stockouts on the same part number, at the same time, in different plants. Plant A froze its parameters in a high-usage year and has quietly built a surplus ever since; Plant B froze in a low-usage year and keeps stocking out on the identical SKU.
Periodic cleanup projects trim the surplus, and then the unchanged parameters rebuild it. A full breakdown of where the spend actually goes shows how much of it traces back to exactly this pattern.
Spot it in your data
Test: For your top 500 SKUs by stock value, compare the lead time stored in the system against actual goods-receipt history for the last two years. Then check when reorder points were last mass-updated.
Decision: If the answer is 'at setup' and lead-time reality has moved, a one-time cleanup will not hold. Commit to continuous recalculation, or accept that the excess regrows on its own schedule.
Root cause 5: MRO procurement is fragmented across plants
Each plant in a multi-site operation typically runs Aquisições de MRO independently: its own requisitions, its own vendor relationships, its own negotiated pricing.
First, one plant raises a purchase order for a part while a sister plant holds surplus of the identical item. Second, multiple plants separately buy the same part from the same supplier at fragmented, non-negotiated pricing.
This is an enterprise-level inefficiency, distinct from any single plant's inventory problem, and it hides the cheapest optimization available anywhere in this chain: the purchase order you cancel before receipt, because the network already had the part. The playbook for fixing this goes into the sequencing in more depth.
Spot it in your data
Test: Take every open purchase order line today and match it against on-hand surplus (stock above maximum, or 12-plus months of cover) for the same or an interchangeable part anywhere in the network.
Decision: Every match is a cancel, a reduction, or a transfer, value captured before the truck arrives, with zero policy change and zero risk. This is usually the fastest first win in the entire chain.
The pattern behind all five
None of the five root causes is a data-entry mistake or a one-off oversight. Each is the predictable output of managing an enterprise-wide, failure-driven MRO supply chain with plant-by-plant tools and once-a-year human review cycles. The fix has to operate continuously and across every site at once, or the gap simply reopens after the next cleanup project ends.
Score your own estate: five questions, five points
One point for each "no" answer. Answer honestly, in one meeting, with no data pull required.
| Question | Point for "no" |
|---|---|
| Can a planner at any plant see, in one search, whether any other site holds a part, regardless of how each site described it? | ☐ |
| Is criticality scored at the part level, not inherited from the asset, and has it been recomputed in the last 12 months? | ☐ |
| Do you know today what share of inventory value has not moved in 24 months, without commissioning a study? | ☐ |
| Have reorder points been mass-recalculated from actual lead-time and demand data in the last year? | ☐ |
| Are open purchase orders checked against network-wide surplus before they are released? | ☐ |
0-1 points: you are the exception; publish how you did it. 2-3 points: you are paying at least two of the four invoices below. 4-5 points: the five tests above will quantify a case your CFO will act on.
See the Five Failure Points in Your Own Extract
Run the five diagnostics above against your own material master, stock, and PO data, guided by a Verdantis solutions engineer.
Obtenha uma demonstração gratuitaContam com a nossa confiança empresas da Fortune 500 e da Global 2000
What the disconnection costs
Four invoices, and you are already paying all of them; they simply never arrive labeled as invoices.
The premium invoice
Emergency buys at a 50-80% premium over standard procurement (MRO360 product documentation), often for parts the network already held elsewhere.
The carrying invoice
25-40% of inventory value sitting excess, obsolete, or duplicated (industry benchmark cited in MRO360 documentation), billing you every year it stays on the shelf.
The fragmentation invoice
Duplicate procurement and non-negotiated pricing across sites, a leak no single plant can see from inside its own storeroom.
The downtime invoice
Unplanned downtime carries an estimated $1.4 trillion annual cost globally, per the Siemens/Senseye O verdadeiro custo do tempo de inatividade 2024 report, a cost a missing spare can trigger directly.
The first three are measurable from your own extracts this week. The fourth is the one your maintenance team already knows by name.
The fix: an AI-native EAM layer that reconnects the chain
The five root causes are not five separate projects. They are one disconnection, expressed five ways, and each maps to a specific, connected capability, not a generic AI claim.
| Causa principal | What closes it |
|---|---|
| Parts invisible across plants | Enterprise-wide search across all sites that normalizes inconsistent naming at query time, so the part is findable regardless of which plant's description it carries |
| Uniform urgency treatment | Part-level criticality scoring, not asset-level, incorporating 20-30 parameters including lead time and substitute availability |
| Dead and obsolete stock | Continuous cross-reference against OEM data and the asset BOM register to flag obsolescence and recommend disposition |
| Static reorder points | Continuously recalculated reorder points, driven by live demand forecasting and lead-time data, so the formula's three variables stay current |
| Fragmented plant-level procurement | Enterprise-wide stock monitoring that recommends an inter-plant transfer before a new purchase order is raised |
Two platform facts matter for anyone evaluating this category. It operates as an overlay above the existing ERP, CMMS, and EAM stack, no system replacement, with approved recommendations writing back into the fields your planners already use. It deploys in 8-12 weeks.
For teams evaluating MRO inventory management software, that overlay architecture is the decisive question to put to any vendor: does the platform require a migration, or does it make the systems you already run smarter?
The platform underneath all five fixes is what Verdantis calls MRO360. A closer look at where the savings actually come from covers the consolidation side of this in more detail.
Who owns each part of this problem
The five root causes surface through four distinct roles, and each role typically only sees its own slice.
| Persona | Feels this root cause first |
|---|---|
| Procurement Planner | Fragmented, duplicate procurement across plants; emergency-buy premiums |
| Inventory Manager | Dead and obsolete stock; carrying cost; locked working capital |
| Planeador de Manutenção | Parts invisible at the point of need; stockouts blocking open work orders |
| Asset Manager / Reliability Engineer | Uniform urgency treatment masking true criticality; downtime risk |
If all four personas report this pain separately, that is itself the diagnosis: the chain is fragmented functionally, and each function is managing its own segment of a problem that only resolves end to end. Work orders and stock finally talking to each other is usually where the Maintenance Planner's version of this gets solved first.
Where this matters most
Everything above is most visible in multi-site, asset-heavy operations, oil and gas, mining, and manufacturing, where the "same part, different plant" failure pattern repeats across dozens of storerooms, and where MRO estates run to tens of thousands of SKUs.
The larger and more distributed the network, the more the five root causes compound rather than simply add together. What this looks like in mining operations specifically is a good next read if that is your setting.
Perguntas frequentes
What is MRO in supply chain?
MRO stands for maintenance, repair and operations: the spare parts, consumables, and supplies that keep industrial assets running, distinct from the direct materials that go into a finished product.
How is the MRO supply chain different from a direct-materials supply chain?
Direct materials follow a production schedule, so demand is forecastable. MRO demand is driven by asset condition and failure, which is why planning rules borrowed from production systematically over-stock some spares and under-protect others.
Why do MRO stockouts and overstock happen at the same time?
Static reorder points get frozen at whatever usage and lead time looked like on setup day. One plant's parameters were set in a high-usage year and now overstock; another plant's were set in a low-usage year and now stock out, on the same part number.
A redução do inventário de MRO aumenta o risco de ruptura de stock?
Only if done without a part-level criticality model first. When critical spares are explicitly identified and shielded, reductions come from over-protected, low-consequence items, and availability of the parts that matter typically improves.
MRO supply chain vs direct-materials supply chain: which needs tighter safety stock?
Neither uniformly. Direct materials need safety stock sized to demand variability; MRO needs safety stock sized to failure consequence and lead time, which is a criticality question, not a volume question.
What data does a plant need before it can diagnose these five root causes?
Standard extracts you already have: material master, stock levels, goods movements, purchasing history, and work orders. A governed material master accelerates the diagnosis, but it starts from extracts, not from a data project.
How long does it take to see results from connecting MRO systems?
An overlay-based deployment typically runs 8-12 weeks to live operation on your own data, since nothing is being replaced. Earlier value often appears in the transaction layer: open purchase orders cancelled once network-wide stock is visible in one place.
Does an MRO overlay platform replace our existing ERP or CMMS?
No. It operates as a governed layer above the existing stack. Recommendations are generated by the platform and write back into the ERP or CMMS fields your planners already use, after human approval.
Is this approach suitable for a multi-site manufacturer with different ERPs per plant?
Yes. The search and criticality layers are designed to normalize inconsistent naming and disparate source systems across plants, which is the specific condition that makes cross-plant visibility hard in the first place.
What is a reasonable first target for working capital release?
20-35% of MRO inventory value on first deployment is a Verdantis deployment benchmark, not a guarantee; the actual figure depends on how fragmented the estate was before the diagnostic.
Who typically sponsors this kind of initiative internally?
Most often a VP of Maintenance, Reliability, or Supply Chain, with Procurement and Inventory Management as co-stakeholders, since the five root causes cut across all three functions.
Can criticality scores be overridden by a maintenance planner who disagrees?
Yes. Every score ships with a written justification, and a planner can override it with their own reasoning. The override becomes part of the model's ongoing training data.
