A material shortage is felt on the line long before it shows up in a report.
A supplier confirms late, or an order simply goes quiet. A substitute might work, but engineering has not looked at it yet. A purchase order date moved and the production schedule did not. A critical part is on the water, sitting in receiving without paperwork, or on a quality hold nobody flagged to planning. Purchasing knows one piece of this, the warehouse knows another, quality knows a third, and the production meeting turns into everyone reading their own screen out loud.
If you run production or planning, the shortage reaches you first and the full picture reaches you last. This guide is about turning supplier and material status into a shared view your whole team works from, instead of a purchasing spreadsheet that only purchasing can see. It is written for the person who has to stand in the production meeting and say which jobs can actually run this week.
What the workflow has to make obvious
A material availability workflow is doing its job when the production meeting stops chasing status and starts making calls. To get there, it has to answer three questions on demand, for any job on the schedule. Which jobs are blocked or at risk right now, and by which material. Whether that supply is genuinely confirmed, or just assumed from an open order nobody has acknowledged. And who owns the next move: expedite, substitute, resequence, or escalate.
When those three answers sit in one place, the shortage conversation shrinks from an hour of phone calls to a few minutes of decisions. The gain here is not a tidier purchasing report. It is that the supplier conversation and the production decision finally sit next to each other, where the person running the line can see both at once.
A test you can run this week
Pick one at-risk job and ask whether your team can see the late material, the supplier's latest promise, the stock on hand, any quality hold, and whether a substitute is approved, without making three phone calls. If they cannot, the current process is hiding the risk until it becomes a stoppage, and by then the room to act has usually gone.
Follow one shortage from supplier email to blocked job
Before changing anything, walk one real shortage through the plant the way it moves today. Not the version in the procedure document. The messy version.
- A supplier email lands moving a promise date, or an order quietly goes past due with no update.
- Purchasing sees it, but the note stays in one person's inbox.
- ERP still shows the original date, so planning keeps assuming the supply is there.
- The part arrives short, or lands on a quality hold, and receiving logs it without telling planning.
- Engineering could approve a substitute, but nobody has told them it is urgent.
- The job hits the schedule, the material is not there, and the line is the last to know.
The expensive part is rarely the shortage itself. It is the days between the supplier's update and the moment production learns that the date it was counting on is gone. Every one of those days was a day you could have expedited, resequenced, or approved a substitute while there was still room. That gap is what the workflow closes.
Map who touches material status, and where it drops
Most of the lost time is not inside any one person's job. It is in the handoffs between them. Purchasing holds the supplier promise. The warehouse holds stock and receiving. Quality holds the incoming holds. Engineering owns the substitute question. Planning holds the schedule and the assumption underneath it. Production lives with whatever falls through. Each of them sees part of the picture clearly and the rest not at all.
Writing that down for one shortage shows exactly where the picture breaks.
| Who | What they see first | What they usually cannot see | Where it breaks |
|---|---|---|---|
| Purchasing team or planner | Supplier emails, confirmations, and promise-date changes | Which specific job the slip threatens | The update stays in an inbox |
| Warehouse and receiving | Stock on hand, what landed, what came up short | Whether an incoming lot is already spoken for by a job | A partial receipt reads as supply |
| Quality | Holds on incoming lots | That a held lot is the only supply for a job next week | A hold nobody passes to planning |
| Engineering | Whether a substitute is technically valid | That the substitute question is urgent | Asked a week too late to use the answer |
| Planning and production | The schedule and the ERP date | Whether that date is confirmed or only assumed | The line finds out when the job releases |
None of these people are doing anything wrong. Each is working from the slice they can see. The workflow's job is to put the slices in one place so the risk is visible before it reaches the schedule.
Sort every critical material: confirmed, at risk, or assumed
The single most useful line this workflow draws is between supply you have confirmed and supply you are only hoping for. Sort every critical material into three states, and keep them honest.
- Confirmed. The supplier acknowledged, the date holds, and the stock is clear of holds. You can plan on it.
- At risk. A late promise, a mismatch, a hold, or a substitute still waiting on engineering. This is the working list for the daily review.
- Assumed. An open order with no acknowledgement. Planning is treating it as real, and it may not be.
Most shortages that blow up a schedule were sitting quietly in the assumed column. Nobody lied; the order was placed, the date was in the system, and everyone treated a request as a commitment. Making the assumed column visible, and chasing those orders down into a confirmed or at-risk state, is often the fastest win in the whole workflow, because it surfaces the shortages you do not yet know you have.
Build the material risk record before adding any automation
The first thing to build is one record per at-risk material, tied to the jobs it can block. A supplier scorecard, a dashboard, and an AI reader can all wait. What earns its keep first is a single place that shows what production can trust, what it cannot, and who is closing the gap.
| Field | What it answers |
|---|---|
| Critical material | Which part, and the jobs or lines it can stop |
| Supplier commitment | The latest promise, and whether it is confirmed or assumed |
| Stock and holds | Quantity on hand, plus any quality hold on incoming lots |
| Substitute | Whether one exists, and whether engineering has approved it |
| Owner and recovery date | Who owns the next action, and when they expect to clear it |
| Evidence link | The email, order, or hold the record is based on |
A purchasing spreadsheet lists what was ordered. A record like this lists what can stop a job and what is being done about it. That difference is what turns a status-chasing meeting into a set of decisions with owners.
Give every at-risk material an owner and a recovery date
A risk list with no owner and no date is just another report to read, and it will quietly stop being read. The moment a material moves into the at-risk state, two fields have to be filled: who owns the next move, and the date they expect the risk to clear. Not the date the supplier promised, but the date the owner will have done something about it, whether that is a confirmation chased, a substitute approved, or a job resequenced.
This is the difference between a red cell and a decision. A red cell says a part is late. An owner and a recovery date say someone is handling it and when you will know more. When every at-risk material carries both, the daily review can move down the list quickly, because the only ones worth discussing are the ones where the recovery date is slipping.
Set the planning window before you widen the net
The first real decision is how far ahead to look. This week's job releases, this month, or the next planning horizon. Look too close and you catch shortages with no room left to act. Look too far and every long-lead order shows up as a risk and the list becomes noise. Start with the window your team already argues about in the production meeting, prove the workflow there, and extend it once the near-term view is trusted.
Connect the least data that keeps the record honest
A working material risk record usually pulls from a handful of places: ERP or MRP orders, the supplier emails or portal replies that update them, inventory and receiving, quality holds, and the engineering approvals behind any substitute. You do not need all of it connected on day one, and trying to integrate everything at once is the surest way to stall. Connect the smallest set that keeps the near-term window current, and add sources only when a real shortage slips through because a source was missing.
Where AI helps, and where it must not
AI earns its place here once the record and the three states are clear, and not a moment before. Its job is to cut the reading and the re-keying, so the facts reach the person deciding sooner. It can read a supplier email and pull out the promised date and the reason for the change. It can reconcile that new date against what ERP still shows and flag the specific job the change puts at risk. It can draft the chase message back to the supplier and the note to planning, ready for someone to check and send.
What it must not do is make the call. It does not approve a substitute, switch a supplier, clear a quality hold, or make the purchasing commitment. Those stay with the people who own them, along with the schedule promise that depends on them. The line to hold is simple: AI reads, reconciles, flags, and drafts; your team commits.
| AI can prepare | Your team decides |
|---|---|
| Pull the promised date and reason out of a supplier email | Whether to accept the new date or push back on it |
| Reconcile the supplier message against the ERP date and flag the mismatch | Whether to expedite, resequence, or hold the affected job |
| Flag which job a slipped date puts at risk | Whether a substitute is acceptable for this order |
| Draft the supplier chase and the note to planning | The purchasing commitment and the schedule promise |
A worked example, invented to show the shape
The scenario below is made up. The company, the parts, and the dates are invented to show how the record behaves under pressure, not to describe a real client.
Picture a mid-size contract manufacturer building assemblies to order. On a Monday, three things are true at once, and until now each lived with a different person. A supplier emails to move an aluminum extrusion out by five days, which threatens a job release the following week. A batch of fasteners has landed but sits on a quality hold, so the line can run only a partial quantity. And an open order for a machined bracket has never been acknowledged, though planning has been treating it as supply for a job in ten days. Dropped into one material risk record, the week looks like this.
| Material risk | Production impact | Evidence | Owner and next decision |
|---|---|---|---|
| Extrusion promised five days late | Job release blocked next week | Supplier email moving the promise date | Planner: expedite, resequence, or ask engineering for a substitute |
| Fasteners on quality hold | Line can run a partial quantity only | Quality hold linked to the received lot | Quality and planning decide the release together |
| Bracket order unconfirmed | Planning is assuming supply nobody acknowledged | Open order with no supplier confirmation | Purchasing team chases confirmation, then updates the state |
None of this is new information to the plant. Purchasing knew about the extrusion, quality knew about the hold, and the bracket order was sitting in ERP the whole time. What changed is that the production meeting now sees all three next to the jobs they threaten, with an owner on each, instead of discovering them one stoppage at a time.
Keep the supplier promise and the schedule in step
A protected material date is only worth something if the schedule reflects it. When a supplier moves a promise and the planner accepts it, the schedule has to move too, or the next shortage is one you created yourself by leaving a stale date in the system. Treat the material risk record and the schedule as two ends of the same decision: the record says what supply you can trust, and the schedule reschedules around it. This is where material availability feeds directly into the production schedule variance workflow, and why the two are usually worth fixing close together.
The first month should make hidden risk visible
Do not try to track every part number at once. Start with the materials that stop jobs, and build outward only when the first version is trusted.
- Pick one critical material group, one supplier, or one planning window to start with.
- Follow a few real shortages from the supplier update to the job they threatened, the way the earlier walk did.
- Agree the three states with the team: confirmed, at risk, assumed.
- Stand up the material risk record with job impact, supplier commitment, stock, hold, substitute, owner, and recovery date.
- Connect the smallest set of ERP, email, inventory, and quality data that keeps it current.
- Add one AI-assisted step, such as reading supplier dates out of email, with a purchasing check on top.
By the end of the month, one material risk review should replace the round of "where is my part" calls, and the orders you were quietly assuming should stop surprising the line.
What tells you it is working
You will feel the change before it shows up in the stoppage numbers. The production meeting spends its time deciding rather than gathering status. Assumed supply gets confirmed or challenged before it reaches the schedule. Substitute and expedite calls happen days earlier, while there is still room to make them. And every at-risk material has a named owner and a recovery date instead of a red cell nobody owns. Fewer line stoppages traced back to material is the result that follows; these are the signs you see first that the workflow is holding.
Common traps
A few failure modes account for most of the value that leaks out of this workflow, and they are worth naming so you can watch for them.
The first is keeping availability inside purchasing when planning, quality, and production all live with the consequences. If only one function can see the picture, the handoffs stay broken. The second is treating an open order as confirmed supply, which is how the assumed column stays invisible until a job stops. The third is discovering the substitute question a week too late to use the answer, when a day's notice to engineering would have kept the option open. The fourth is a risk list with no owner and no recovery date, which decays into a report nobody opens. The fifth is trying to track every part number instead of the handful that actually stop jobs, which buries the real risks under noise and guarantees the list falls out of date.
How Ubisar would build this with you
In the first week we would pick one critical material group or one supplier with you and trace a few real shortages from the supplier update through to the job at risk, the substitute question, and the expedite or resequence call. The first thing you would have is a material risk record that shows job impact, supplier commitment, stock and hold status, the substitute option, an owner, and a recovery date, each linked back to the email, order, or hold it came from.
Over the next couple of weeks we connect the least ERP, email, inventory, and quality data needed to keep that record honest, and add AI only where it clears manual reading, pulling dates out of supplier mail and drafting the chase, while the substitute, supplier, and quality decisions stay with the people who own them. If your team tends to learn about shortages when the line goes quiet, the place to start is a single review of one material group. You can see how a month of that work runs in the AI, Data & Tech Implementation Service, or tell us the workflow that keeps stopping your line on the contact page.