When the Inspection Fails: Rework, Sorting and Seconds in Knitwear

By YZR Sweaters · 2026-10

The call every buyer dreads: the inspector has rejected the lot and the vessel cut-off is four days away. In our experience the outcome is decided less by the defect itself than by how quickly the buyer sorts the problem into the right bucket. Some failures are genuinely recoverable in a week. Others cannot be fixed at any price, and the only real question is who absorbs them. Here is how we work through it with customers.

Read the report before you react

Ask three questions of the inspection report before anything else:

• What is the defect classification? Critical, major and minor are not the same conversation. A lot rejected on critical defects — needles, broken glass risk, choking hazards on children's knit — does not get reworked, it gets controlled. Our AQL guide covers how the classifications work.
• Is the failure systemic or random? If every piece in the sample has the same measurement drift, that is a pattern or a machine setting and it affects the whole lot. If three pieces out of eighty have unrelated faults, that is process noise and sorting may be enough.
• Is it one defect or several? A lot fails on a combination surprisingly often — a minor measurement issue plus a shade issue plus loose threads. Each needs a separate decision, and combining them into one panic usually leads to a bad outcome.

Category one: what can genuinely be reworked

These are the faults that a decent factory can put right without changing the garment's character:

• Loose threads and untrimmed ends. The most common reason a lot fails, and the cheapest to fix. Hand-finishing teams can clear a few thousand pieces in a couple of days.
• Pressing and shaping faults. Creased collars, uneven blocking, poor placket pressing. Re-pressing is straightforward and usually improves the garment rather than compromising it.
• Soiling and handling marks. Often from packing rather than production. Spot cleaning or re-washing may fix it — but check whether re-washing changes the hand feel before approving it.
• Minor linking and seam faults. A dropped stitch at a shoulder or an imperfect link can often be re-linked by hand. Our notes on linking quality explain what a good link looks like, so you can judge whether a repair will be visible.
• Label and trim faults. Wrong placement, crooked care labels, missing hangtags. Tedious but fixable, and a strong candidate for rework at the packing stage.

Category two: what can only be sorted

Some problems cannot be fixed, but the good pieces are still good. This is where 100% sorting comes in:

• Shade variation within a lot. If the dye lot drifted, individual pieces may still fall inside tolerance. Sorting by shade into acceptable and unacceptable piles works — but agree the reference shade first, ideally against the approved colour card or a retained bulk swatch.
• Measurement spread. If the pattern was slightly off, some sizes may be in tolerance and others not. Measure, sort by size, and see what you actually have before you write off the lot.
• Isolated surface faults. A snag here, a dropped stitch there. If the incidence is low enough, sorting removes them.

Two things to get right. First, write a sorting criteria sheet — a written definition of what passes, ideally with photos, because a sorting team working from verbal instructions will drift within a day. Second, price the sorting honestly: it is charged per piece, it takes days, and it has to be done by people who understand knitwear, not by whoever is free.

Category three: what cannot be fixed

This is the bucket buyers hope they are not in, and it is worth recognising early:

• Wrong yarn or unapproved substitution. If the composition or the yarn grade is not what was approved, no amount of rework changes it. This is a contract issue, not a QC issue — see our guide to yarn substitution.
• Systemic shrinkage or dimensional instability. If the garment moves beyond tolerance in washing, re-washing will not stabilise it. The construction or the yarn is wrong.
• Colourfastness or RSL failures. A test failure cannot be reworked out. The goods cannot be sold as specified.
• Wrong gauge, weight or hand feel. The fabric is what it is. This is why wash testing and grammage approval on the pre-production sample matter so much.
• Pilling propensity. A fabric that pills badly will keep pilling. Returns will follow it into the market.

When you are in this bucket, the conversation shifts to commercial remedy: partial credit, replacement production, or accepting the goods at a discount for a channel that can carry them.

Timing: where the failure lands changes everything

The same defect has very different consequences depending on when it is found:

• Before packing. Best case. Rework can happen on the production line with the original people.
• After packing, before vessel cut-off. Workable but expensive — cartons have to be opened, reworked, re-packed, and the booking may need moving.
• After the vessel sails. Now you are negotiating about goods on the water. Your options are port-side rework, a destination-country sorting operation, or a claim. All three cost more than fixing it at the factory.
• After delivery to your DC. Worst case, and the reason in-line inspection at 20–30% of production is worth the money. Our guide to pre-shipment inspection covers the timing question in more detail.

Seconds: what to do with them

Sorted-out pieces and reworked lots produce stock that should not go into your main channel. Decide what happens to it before it exists:

• Off-channel sale. Outlet, employee sale or a discount channel that is clearly separated from your main retail. Works if the fault is cosmetic and the customer knows what they are buying.
• Negotiated discount on the whole lot. Sometimes cleaner than sorting: accept everything at an agreed reduction rather than paying for sorting.
• Destruction. For branded goods where the risk of seconds appearing in the wrong channel outweighs the recovery value. Get it documented with photos and a certificate of destruction.
• Never the quiet option. Shipping seconds into your main distribution without telling anyone is how a brand loses a retail account. If the fault is invisible to the customer and genuinely cosmetic, say so in writing — but say it.

Our notes on dead stock clearance cover the commercial side of moving problem inventory.

What rework does to the garment

Worth being honest with yourself about: reworked goods are not identical to goods that were right first time.

• Re-washing changes hand feel. A second wash cycle softens or relaxes the fabric, so a re-washed lot can feel different from the approved sample.
• Re-pressing can flatten surface texture. Ribs and cables lose definition with repeated pressing.
• Hand-linked repairs are visible to a trained eye even when they are acceptable.
• Re-packing loses time. Cartons, polybags and labels may need replacing, which adds cost and can push the booking.

None of this means rework is wrong — it usually is right. It means you should see a reworked piece before you accept the lot, not just a written confirmation that rework is complete.

Who pays

This is settled by what was agreed before production, which is why it belongs in the contract rather than in the argument:

• Was an AQL specified in writing? If no standard was agreed, "the goods aren't good enough" is a matter of opinion. Specify the standard and the defect classification up front — see quality standards documents.
• Was there a remediation clause? Who pays for rework, for sorting, for re-inspection, and for freight if the booking moves. Our overview of contract essentials lists the clauses worth having.
• Was there an approved counter sample? Most disputes come down to "that is not what I approved". A retained, signed pre-production sample settles it quickly.
• Payment structure protects you. If the balance is payable before shipment, you have leverage; if it is already paid, you are negotiating from behind. See payment terms.
• Chargebacks work both ways. Large retailers will charge back for quality failures, so think about what your own customer's standard does to your exposure — the discipline in complaints and returns handling applies here too.

Preventing the next one

Every failed lot has a root cause, and in our experience it is usually one of four things:

• A gap in the spec. Something was never specified, so the factory filled the gap sensibly and differently. Fix it in the tech pack.
• A change that was not costed through. Late changes to yarn, trim or construction are the single biggest source of bulk-vs-sample differences. Our notes on order changes explain how to handle them.
• Rush. Compressed production skips steps. The problems show up at inspection or, worse, in the market.
• No in-process control. A pre-production meeting and an in-line check at 20–30% catch most problems while they are still cheap. See the pre-production meeting.

YZR Sweaters works with BSCI, SEDEX and Oeko-Tex certified partner factories in Dongguan and supports inspection by SGS, BV or your own nominated inspector. We keep a retained pre-production sample on every style so there is never an argument about what was approved, and we tell you when a lot fails before you hear it from a third party. Where rework is the right answer we scope it, cost it and re-inspect before anything ships.

Frequently asked questions