What Do You Do When a Reorder Arrives Rejectable?

The first hour after a bad reorder lands decides most of what follows. If it is spent describing the problem, the next week is spent arguing about it. If it is spent measuring, the next week is spent choosing between four remedies that already have costs attached.

The distinction that matters, and it should be settled before anything else, is whether the lot contains defective units or whether the lot has moved. Those are different findings with different remedies, and treating the second as the first is how a sorting operation costs a month and fixes nothing.

It also decides who is right, which is worth saying because most disputes are not about money. A buyer who can point to a band and a reading is asking a factual question; one who cannot is asking for a concession, and the two are answered differently even when the underlying problem is identical.

What Makes a Reorder Rejectable Rather Than Merely Different?

Three findings. A property outside a written band. A registered component that is not the registered component. Or a defect count above the reject number for the sample size. Anything else is a difference, and a difference without a band is a conversation rather than a rejection.

That is uncomfortable but it is the truth of the situation: without a written band, a retained reference and a sampling plan, there is no standard to reject against. The remedy then is not commercial but procedural - write the standard, then judge the next run against it.

The answer to the fourth question - whether the previous run shows the same finding - is the one that separates drift from a one-off. If the retained unit from the previous run measures the same way, the programme has been shipping this for two runs and the conversation is very different from the one about a single bad lot.

How Is the Finding Established Before the Conversation?

Four questions, answered in writing before anyone discusses money. Which property, and by how much. On how many units, and from which part of the run. Against which reference, with which method. And whether the same finding appears on the retained unit from the previous run.

There is a fifth option that is easy to forget because it is not a remedy at all: accept the lot and revise the reference. Where the difference is consistent, acceptable to the market and less costly than any correction, updating the file is more honest than a concession dressed as a settlement.

What Are the Four Remedies and When Does Each Apply?

Sort and rework, where the defects are concentrated and separable. A one-run exception, where the difference is visible, the season cannot wait, and the buyer accepts it in writing. Replacement, where the lot cannot be corrected. And a credit, where the goods are usable but not what was paid for.

The order matters because the cost rises sharply down the list. A sort costs hours. An exception costs a concession on the next order. A replacement costs a production slot and calendar time. A credit costs the margin on the run.

One caution about rework on a reorder: it rarely restores a bag to the approved standard. A resewn seam sits differently from an original one, and a reworked panel carries the marks of the repair. Rework is appropriate where the difference is functional and inappropriate where it is visible.

How Is Rework Specified and Checked?

As a written instruction with a method, a standard and a re-check: what is to be done, to what standard, on which units, by whom, and how the result is verified. Five fields, and the re-check is the one most often left out.

Scope

Which units, identified by carton and by count

Operation

What is to be done, in enough detail to be repeated

Standard

The reference the reworked unit must meet

Method

How the rework is carried out, and with which tools or materials

Re-check

What is inspected afterwards, and to which sampling plan

Rework also needs its own record. A reworked lot has two histories - the original run and the rework - and blending them into one sheet is how a programme loses the ability to say which units were touched and when.

When Is a Credit the Right Answer?

Where the goods are saleable and the difference is real. A shade that sits just outside the band, a hardware finish that is slightly off, a dimension within tolerance but at the edge: all are cases where the product works and the buyer has nonetheless not received what was specified.

A credit is the remedy that most often goes unrecorded, because it is settled in an email about money. It should be recorded as a finding with a remedy, on the same log, because otherwise the pattern of small credits never becomes visible as a pattern.

Where the replacement is urgent, ask what can be done in parallel rather than in series. Booking the material at the same time as the slot, rather than after the decision, is the difference between a replacement in one window and one in two.

What Does Replacement Involve on a Repeat?

A production slot, and that is the expensive part. Bulk production is 35-50 days from written sample approval, so a replacement run costs that window plus whatever waiting time precedes it. Where the season cannot absorb it, the realistic options narrow to an exception or a credit.

This is why the finding should be established at inspection rather than at the destination. A lot found before it ships can be reworked in the factory in a week. The same lot found after arrival has to be reworked at the destination, or replaced, and both are far more expensive than the inspection would have been.

Agree the allocation rule in the contract rather than after the event. A rule written before a failure is a clause; the same rule proposed after one is a claim, and it is read very differently by the person paying.

How Is the Cost of a Failure Allocated?

By cause, and this is where the record earns its keep. Where the failure traces to an unrecorded substitution, the cause is on the production side. Where it traces to a band that was never achievable, the cause is shared. Where it traces to a buyer change that never reached the file, the cause is on the buyer side.

Keep the count in units as well as in value. A write-off recorded only as a sum cannot be compared with a production figure, and the comparison is what tells a programme whether a failure affected two per cent of a run or twenty.

What Happens to the Units That Cannot Be Sold?

Four options, in ascending order of difficulty: sell them through a channel where the difference does not matter, rework them, hold them for a future season against a revised reference, or destroy them. Where a product carries a brand or a tested claim, the last one is sometimes the only honest answer.

Whichever applies, record the count and the disposition. Units that were made, rejected and quietly forgotten are the origin of the question that arrives two years later about whether a batch of old stock can be shipped against a current order.

Where the same finding has appeared before, say so on the record. A second occurrence of a finding that was already corrected is a different problem from a first one, and it usually means the correction was documented rather than actually applied.

How Is the Outcome Written Into the File?

Three lines: the finding, the remedy, and the correction that prevents recurrence. The third is the one that matters and the one most often omitted, because by the time the commercial settlement is agreed everyone wants the matter closed.

One more change is worth making after any failure: shorten the interval to the next comparison. A programme that has just had a run fail should check the following run against the approved one rather than only against the previous one, because that is the check that shows whether the correction held.

What Changes About the Next Run?

One of four things: a register line gets named properly, a band gets reset on evidence, a check gets added to the routine sheet, or a material gets a comparison before it is cut. If none of those happens, the same failure will recur, usually within two runs.

And one more thing should change: the inspection point. A failure discovered at the destination is a failure that should have been discovered at the factory, and moving the comparison earlier is the single change that prevents the expensive version of the same problem next time.

Finally, close the matter in writing even when the settlement was verbal. A file that records the finding, the remedy and the correction is what lets the next person answer, a year later, whether this programme has had this problem before.

A final practical point: agree who owns the units while the matter is open. Goods in dispute that have already shipped, already been paid for, or already been unpacked are each a different commercial situation, and saying which applies early prevents the argument that otherwise follows.

One more thing belongs in the file: the date the matter was closed. A finding that stays open across two bookings is a finding that will affect the next run, and a closed date is what stops it being treated as settled when it was only deferred.

Frequently Asked Questions

People Also Ask