What you actually own in this category
Sourcing decides who supplies the magnet and at what price. Engineering decides which magnet. The materials manager owns everything in between and after: whether it is on the shelf when the line calls for it, whether it can be traced when quality asks, whether it is still usable in eighteen months, and whether storing it created a hazard nobody wrote down.
That job is harder for magnets than for most purchased parts, for reasons that have nothing to do with buying them well.
None of these are exotic. They are simply not the defaults your planning system assumes, which means each one has to be handled deliberately rather than inherited from the parameters used on fasteners.
Setting MRP parameters for a long-lead, volatile part
Most magnet availability problems trace back to planning parameters that were set once, from a quoted lead time, and never revisited. Four fields do most of the damage.
| Parameter | Common setting | Better practice for magnets |
|---|---|---|
| Planning lead time | Supplier's quoted lead time | Quoted time plus observed variance, plus a customs and documentation buffer |
| Lot size | EOQ from carrying cost | Reconciled against MOQ and tooling amortization, which usually dominate EOQ |
| Order multiple | 1 | Supplier's tray, tube or pack quantity — broken packs invite handling damage |
| Time fence | Short or none | Firm zone covering at least the manufacturing lead time, to stop MRP churn |
| Reorder point | Average demand over lead time | Demand over lead time plus safety stock sized on lead-time variance |
| Source of supply | Single default vendor | Split-source rule where a qualified alternate exists, so it stays warm |
The time fence is the one to fix first. Magnets typically sit far enough out in the lead-time horizon that ordinary demand noise reschedules them constantly. Every reschedule the supplier receives makes your forecast less credible, and a supplier who has stopped believing your dates will not reserve material against them. A firm zone that covers the manufacturing lead time converts a stream of contradictory signals into a commitment the supplier can actually plan against.
Record the quoted lead time on every RFQ and plot it by supplier and grade. A drift from six weeks to ten over two quarters is the earliest reliable signal of tightening supply, it appears months before delivery performance degrades, and it is sitting in data you already generate. Feed it back into the planning parameter rather than leaving the original number in place.
Sizing safety stock on the right variable
The standard safety stock calculation accounts for variability in both demand and lead time:
// z = service factor, L̄ = mean lead time, d̄ = mean demand
// σd = demand variability, σL = lead-time variability
For most purchased parts the demand term dominates. For magnets it is usually the lead-time term, because σL is large and is multiplied by mean demand. The practical consequence is that better forecasting yields much less improvement than reducing lead-time variability — which is a supplier and contract problem, not a planning problem.
That formula also covers only ordinary variation. It says nothing about a supplier being unable to ship at all, which is the failure mode this category actually presents. Critical parts need a second, separate layer:
// hold this only where no warm second source exists
// time to alternate includes any customer or regulatory approval
Sizing disruption cover on requalification time rather than on part value produces results that look wrong on a stock report and are right in practice. A twelve-cent magnet with a nine-month qualification path may warrant more months of cover than a part costing forty times as much with three approved sources. Classify by consequence, not by unit cost.
Consignment and VMI shift who finances the inventory. They do not shift who suffers if the supplier cannot obtain material, and a consignment agreement with no minimum on-site quantity can leave you holding nothing at exactly the wrong moment. Specify the floor quantity, the replenishment trigger and the notice period if the supplier wants to reduce the position. See blanket POs, consignment and VMI for the structures.
Receiving, lot control and traceability
Magnetic properties vary between production lots within the allowed tolerance band. That variation is normal and acceptable, and it is exactly why lots must stay separated. Once two lots are commingled in a bin, you cannot answer a containment question without quarantining everything.
Chasing certificates after the fact is the most common avoidable failure here. If documentation is not captured at receipt, it will be needed on a day when the supplier is slow to respond and the part is on a customer hold. Make the certificate a receiving requirement rather than a follow-up task, and keep it against the lot — see certificates and compliance docs for what each document actually covers and incoming inspection for sampling.
Where magnets feed a PPAP or regulated product, traceability may need to extend past the magnet lot to the alloy batch behind it. That is a question to settle during qualification, not during an audit, because not every supplier can provide it and the ones who can may need notice.
Storing and handling magnetized material
This is the part of the job with no equivalent in other categories, and the part most likely to be handled informally until something goes wrong.
Field strength falls off steeply with distance, which is reassuring for a single part and misleading for a pallet. Stacked and palletized magnetized stock produces a combined field far larger than any individual piece, and stores layouts are usually planned around a sample on a bench rather than around the quantity that actually arrives.
Two of these are worth emphasizing because they produce damage that is invisible until much later. Storing above the rated temperature causes irreversible flux loss, and the parts will look perfect — the loss only appears when the assembly underperforms in service. Repeated condensation on coated NdFeB attacks any chipped edge, and corrosion that begins under the coating spreads before it is visible.
If you move stock between plants, review shipping magnetized material before the first transfer rather than after a shipment is refused. Packaging that is compliant for one mode is not automatically compliant for another, and the shielding requirement is a real constraint on how inter-site transfers get packed. For the thermal limits behind the storage rule, see magnets and temperature.
Shelf life, obsolescence and engineering change
Sintered NdFeB does not meaningfully lose magnetization sitting on a shelf at room temperature. Left within its rated temperature and protected from moisture, a magnet stored for years will perform as it did on arrival. The shelf-life risk in this category is not magnetic — it is the coating, the packaging and the drawing revision.
Engineering change is where materials managers earn their keep in this category, and where they are most often consulted too late. A grade change approved on Friday can strand a year of inventory that was bought against a long lead time and a large MOQ — a combination that guarantees the stranded quantity is large.
Put materials review into the change process
Before any magnet grade, coating or dimensional change is released, the change record should carry current on-hand quantity, on-order quantity, weeks of cover, and the disposition plan for both. That is a five-minute lookup that routinely prevents a five-figure write-off, and it only happens if the process requires it rather than depending on someone remembering to ask.
Prefer run-out to cutover
Where the change is an improvement rather than a correction, an effective-on-depletion date lets existing stock be consumed. Where it is a genuine correction, quarantine the balance immediately and decide disposition explicitly — unresolved stock quietly ages in a bin until it becomes an audit finding.
Rare-earth magnets contain recoverable neodymium, praseodymium and, in higher grades, dysprosium or terbium. Recycling infrastructure has expanded specifically to recover this material. Before disposing of stranded inventory, ask your supplier what recovery route exists — and record magnet chemistry and mass in the item master now, because that is the information a recovery route will ask for.
The monthly materials review
Keep the review short and focused on the measures that actually predict a shortage, rather than on the ones that explain the last one.
| Measure | Why it matters here | Action trigger |
|---|---|---|
| Weeks of cover vs time to alternate | The only cover ratio that reflects real risk | Cover below the requalification window on any critical part |
| Lead-time variance by supplier | Dominant term in safety stock for this category | Variance widening two periods running |
| Quoted lead-time trend | Earliest available signal of market tightening | Any sustained upward drift |
| Lot traceability completeness | Determines containment cost if quality escapes | Any receipt without a matched certificate |
| Aged and slow-moving stock | Coating and packaging degrade before magnetics do | Stock past its FIFO window or in damaged packaging |
| Open engineering changes | Single largest obsolescence driver | Any magnet change without a disposition plan |
| Supplier on-time delivery | Confirms whether parameters match reality | Two consecutive misses on the same part |
Cycle counting deserves a note of its own. Magnets are small, dense, easy to miscount and prone to clumping in a bin, and count accuracy in this category tends to be worse than the site average while appearing fine on a weight-based check. Where parts are counted by weight, verify the piece weight against the drawing periodically — a coating change or a dimensional revision alters it, and the scale will keep reporting confident, wrong numbers.
Where this connects outward: cover targets should be set from the part-level scoring in supply risk monitoring, and the inventory posture for each part family should follow the sourcing model set in the category strategy. Keeping the numbers in one place stops planning and sourcing from working to different assumptions about the same part.
Weeks of cover measured against time-to-qualified-alternate, for your critical magnet part numbers. It is the measure that tells you which shortages you could actually recover from, it is understood outside the materials function, and it is the argument that gets inventory approved before a disruption rather than after one.