← Manufacturing collection
Manufacturing Maintenance Business Central

Spare Parts Inventory Management Fails When the Low-Stock Flag Has No Owner

A technician says the part is low. Nobody owns turning that sentence into a purchase order, so one plant hired a person to be the missing owner. Where the step breaks, what Business Central can see by item type, and how to time the gap on your own floor.

Christopher Wakare
Updated
7 min read
Manufacturing

One toolmaker described the failure on X this month, about EDM wear parts that are always backordered: "You'll use it, mention it to the boss and it never got reordered. Then next time it goes down you are in a pickle. You end up paying 3 times the amount to get it from a 3rd party in a rush." A technician saw the part running low and said so out loud. Spare parts inventory management in a maintenance department fails one step later, between the person who saw the bin and the person who can place the order.

A parts coordinator posted the other half of the same problem: "I was hired in my position to avoid this issue. Tech tells management we were out of stock or low stock, and it would get dropped." The job covers 40 technicians across four geographic locations. When I replied, I asked one question: how long does it take from a technician's low-stock signal to the person who can actually place or expedite the order, across those four locations? Both posts are anonymous, so I use them as color, not as data.

The definition

Spare parts inventory management is the work of having the right parts on hand when a machine goes down: wear parts, critical spares and consumables held in the maintenance storeroom, also called MRO inventory management. It has a stocking half (what to hold, how many, where) and an ordering half (who turns a low-stock signal into a purchase order). This article is about the second half.

How this was sourced

Analysis, not a deployment write-up; no customer deployment is described here. The two anonymous X posts (November 2026) are color, not evidence. The documented seam rests on a Plant Services panel report (7 April 2026), a Plant Services podcast (5 March 2026) and four ReliaMag pieces, three of them trade articles by consultants and trainers and one an editorial statistics guide, each linked where cited. The Business Central statements come from Microsoft Learn pages read on 14 November 2026.

The seam is documented outside those two posts. Ricky Smith, CMRP, writes in ReliaMag (undated) that "in many organizations, the storeroom reports to procurement or finance, not maintenance," and that procurement's and maintenance's goals "conflict constantly when nobody's mediating them." Phantom inventory, he says, "creates emergency purchases at premium prices." Joe Kuhn, a former plant manager now running a reliability consultancy, described the floor end on a Plant Services podcast (5 March 2026): crews "don't have the right part, and they really don't do any productive work until about 10."

A spare parts audit starts at the shelf: count accuracy, criticality rankings, min/max values. That audit assumes somebody owns the step after the flag. In the two posts above the shelf was empty, a person said so, and no order followed. A wrong min/max is a different failure, covered in the Business Central reorder point accuracy article. Here the signal is right and nobody acts on it.

Who manages spare parts inventory and procurement on your floor?

On paper, two people hold a spare part. The storeroom lead or maintenance manager holds the stock, and a buyer in purchasing places the order. Andrew Gager, CEO of AMG International Consulting, describes a client moving MRO storeroom oversight from Finance to the Maintenance Manager (ReliaMag, 30 June 2026) and tells plants to "Lock the storeroom and assign clear accountability." He calls stockouts on critical spares "reliability failures wearing a purchasing department uniform." Smith's fix in the storeroom article is a reporting line to the maintenance manager or "a dedicated warehouse/inventory control position that sits within the maintenance organization." Those assignments cover the stock and the purchase. Neither covers the sentence a technician says at the bench.

Ricky Smith, CMRP, writing on wrench time for ReliaMag (undated), names who fills that vacancy: without planning, the technician "becomes the planner, the parts chaser, and the permit coordinator." The toolmaker's version is shorter. The technician tells the boss and carries the memory of it alone. At the 2026 Fluke Xcelerate panel that Plant Services reported on 7 April 2026, Jason Hahn of Amsted Automotive named a stockout "on some MRO item" last week as a root cause his team has to analyze like any machine failure.

The coordinator is the plant's answer to the same vacancy, filled by a hire. As I read the post, the role is four steps: receive the flag, check stock across locations, place or expedite the order, close the loop. Spread over 40 technicians and four locations, that is a job "almost entirely dedicated to making sure everyone has what they need," in the poster's words.

Maintenance spare parts management runs on a conversation

The flag travels as a sentence: said to a boss, passed at a shift handoff, typed into a chat. A sentence has no due date and no owner, and Business Central cannot see it. Whether Business Central can see the part at all depends on one field. Microsoft Learn describes the Non-Inventory item type as "physical things, such as bolts or pens, that your business consumes but doesn't fully track in inventory," and states that "The Service and Non-Inventory types don't let you track inventory quantities and values" (Microsoft Learn, Understand item types, updated April 2026). Its feature table marks Physical Counting and Planning as No for Non-Inventory items, against Yes for Inventory items.

A wear part set up that way has no on-hand quantity. No threshold can cross, no automatic planning run can propose an order, and no alert layer can flag it low. The only record that it is low is the technician's memory, which is the failure the toolmaker described.

If your plant tracks parts requests or on-hand counts in a CMMS, the same gap sits one system over: a low-stock request can exist as a CMMS record or work order and still never become a Business Central purchase order. OpsGrid reads Business Central only, so it does not see a CMMS record either, and I make no claim about any CMMS integration.

Setting the part up as an Inventory item does not finish the job. Learn says "To include an item or SKU in the planning process, you must assign it a reordering policy. Otherwise, it must be planned manually" (Microsoft Learn, Design details: planning parameters, page updated March 2026), for example on the Order Planning page. An item with a quantity and no policy waits for a person to look.

My practice rule, which Microsoft does not state: set a wear part or spare up as an Inventory item when it is critical, expensive, slow to arrive, held in the storeroom or able to stop a machine. Keep Non-Inventory for low-value, readily available consumables you expense when you buy them; the same Learn table marks Location Transfer as No for that type, so a part you may need to move between locations cannot be one.

Two consequences come with that rule. First, the quantity is only as good as the count. Smith names the failure at the bench: "The tech shows up to pull the part, and the bin is empty." Gager's first move is a physical count of the top 50 critical spares, because a planner "builds a full week of work around parts the count claims exist, then the crew opens the bin and finds dust" (ReliaMag, 30 June 2026). A wrong count is a third failure beside the missing owner and the wrong min/max. Second, the controller signs off on the change. Microsoft Learn records non-inventory cost in "Cost Amount (Non-Invtbl.)" and says "Costs for these types of items aren't reconciled to the general ledger." Moving wear parts to Inventory puts them into inventory value and the physical count cycle, which finance owns.

The spare parts ordering process has one step nobody owns

Written down, the ordering process is five steps. The two posts break it at the first two and at the last.

Step What has to be true Owner to name Where the posts show it breaking
1. Flag The technician's "low" becomes a record in the same shift The technician records it; a named role receives it "Mention it to the boss": a sentence, no record
2. Receive One person per location owns the inbox for flags, with a named delegate Parts coordinator or storeroom lead The boss hears it and it goes no further
3. Check stock On hand at every location and open purchase orders, checked before anyone orders The same owner A Non-Inventory part has no quantity to check
4. Order or expedite A person with purchasing authority places or expedites it by an agreed date A named buyer per location Not shown in the posts. This is the interval my question asks about
5. Close The flag closes only when a purchase order number sits on it and the part is on the shelf The owner from step 2 "It would get dropped"

Step 5 is where the coordinator's phrase lands. A flag with no close condition stays open in somebody's head until the machine goes down. A low-stock flag that lives in a sentence has no due date.

MRO spare parts management math: the rush buy and the full-time stocker

The toolmaker's three times is one figure for one part at one shop. The downtime benchmark most often repeated has no primary source behind it either: ReliaMag's MRO and spare parts statistics guide (undated) traced the claim that 50 percent of unscheduled downtime comes from spare parts shortages to a June 2018 IBM press release that cites Aberdeen Group with no report, year or method, and found no locatable Aberdeen publication containing it. Take your numbers from your own purchase ledger.

Price the choice on three separate lines. Cash is the premium on each rush purchase order: the planned price against the price you paid a third party, summed over a year. If the toolmaker's multiple held on your part, each rush buy would cost two extra planned prices. Capacity is the technician hours spent chasing parts, plus the coordinator's loaded cost if you hire one. Upside is the hours the machine stood, times the contribution margin per hour you lost. If you are the Plant Head or VP Operations at a discrete manufacturer of 100 to 3,000 employees on Business Central, these are the three lines you take upward. Do not add the three into one number. A hire justified on avoided downtime needs different proof from a hire justified on rush premiums, and the board should see each line by itself.

The post gives no salary for the coordinator and I will not guess one. The key-person question is easier to ask. The hire works while that person is in the building and reading every flag. When they take leave, the order step reverts to the boss's memory. The key-man risk article covers the same pattern for operating knowledge.

Wear parts versus critical spare parts

Gager sorts spares into three tiers: critical, essential and commodity. ReliaMag's statistics guide adds that critical stock levels follow failure consequence and supplier lead time, not a universal ratio. That sorting handles critical spares. For a critical spare, the decision is whether to hold one, a capital call with finance in the room. A wear part is a different case. In my reading, it wears out on a cadence, so the need is predictable, and a predictable need is the one nobody treats as an event. The toolmaker's parts are always backordered, a supplier lead-time problem that the gap between purchase order date and posted receipt in Business Central will show. For a wear part, the decision is who orders the next one and by when.

Where a system can act on the flag

In decision infrastructure terms, the missing piece is an owner and an execution path for the signal: a named person receives it, and the approved action reaches the ERP without rekeying. What a system can do depends on the item type.

For Inventory items with a reorder threshold in Business Central, the signal already exists in BC. The live reorder-threshold signal in OpsGrid (in beta) checks it on a schedule and pings the owner in Teams. On Approve it creates a Draft PO in Business Central, which a buyer opens, confirms and submits. That loop starts from BC's number. A part set up as Non-Inventory has no number, so there is no signal for it to act on. OpsGrid's live signals read Business Central only, so a flag that exists only as a sentence, or only as a CMMS request, sits outside them.

Free, 3 minutes

The Decision Latency Diagnostic is a 12-question self-score of where time leaks between a signal and the action that follows. Run it with your last rush buy in mind.

Frequently asked questions

Who manages spare parts inventory and procurement?

Two people hold a spare part on paper: the maintenance manager or storeroom lead holds the stock, and a buyer in purchasing places the order. Where the step between them has no owner, a technician says a part is low and nothing follows. Someone has to record the flag, check stock across locations, and order or expedite. Name one person for that step at each location, with a named delegate, and set a date by which a flag must have an order behind it.

How do you manage spare parts inventory?

Start with the order step, then the shelf. Name who receives a low-stock flag and who can place or expedite the order. Set critical and wear parts up as Inventory-type items in Business Central so the system holds an on-hand quantity, and give each a reordering policy if you want planning to include it (Microsoft Learn says an item without one must be planned manually). The Inventory rule is the author's practice view, not a Microsoft requirement. Tier parts by what a stockout does and how long the supplier takes. Then time the gap from first flag to purchase order on your last three rush buys.

How do you determine critical spare parts?

Rank each part by what its absence does to the machine and by how long the supplier takes to deliver it. ReliaMag's MRO statistics guide (undated) says critical spare stock levels follow failure consequence and supplier lead time, not a universal ratio. Andrew Gager, writing in ReliaMag (30 June 2026), sorts spares into critical, essential and commodity and recommends counting the top 50 critical spares first. A wear part that fails on a cadence is a different case: for it, the question is who orders the next one and by when.

Does Business Central have purchase requisitions?

Business Central's purchasing overview (Microsoft Learn, updated June 2026) lists purchase quotes, invoices, orders, blanket orders, credit memos and returns, and describes no requisition document. It does have a Requisition Worksheet, which the planning documentation (Microsoft Learn, updated June 2026) says lists items you want to order, filled in manually or by the Calculate Plan batch job. That is a planner's worksheet, not a request a technician submits from the floor, and the cited pages describe no such step. Any purchase process can sit under an approval workflow.

Pull your last three third-party rush buys. For each, find the date a technician first said the part was low and the purchase order date in Business Central. The system can give you the second date. If nobody can give you the first, the flag lives in a sentence, and that is the finding.

How long does a low-stock flag live in a sentence on your floor?

The Decision Latency Diagnostic scores where time leaks between a signal and the action that follows: 12 questions, 3 minutes. Run it against your last rush buy.

Take the diagnostic →

The Execution Edge

Monthly. For operations leaders building faster on AI. Real case studies, system blueprints, and tools, no fluff.

Your subscription could not be saved. Please try again.
Your subscription has been successful.