Your Pre-Shipment Inspection Failed — Now What?

A FAILED result on a pre-shipment inspection report feels like a disaster. Your production is finished, your ship date is booked, and a customer — or a marketplace deadline — is waiting on the other end.

Here is the reframe that matters: a failed inspection is the system working. The defective goods are still in the factory, not in your warehouse — or worse, in your customers' hands. The few hundred dollars you spent on the inspection just saved you from paying freight, duties and returns on a shipment you would have rejected anyway. What happens in the next few days determines whether this becomes a two-week delay or a five-figure loss.

This guide walks through exactly what to do, in order.

First, understand what "FAILED" actually means

A professional inspection does not fail a shipment on opinion. It fails on numbers.

Your inspection ran against an AQL sampling plan (usually ISO 2859-1 / ANSI Z1.4): a sample was drawn — say 200 units from a 5,000-unit order — and defects were counted by class. Each class has an accept number (Ac) and a reject number (Re). If the defects found in any class reach the reject number, the lot fails that class. That is the entire test.

So before anything else, open the report and find:

• Which class failed — critical, major, or minor. A critical fail (one unit that could hurt someone or breach regulations) is a different situation from a minor fail (too many cosmetic flaws). See critical, major and minor defects explained.
• By how much — 8 majors against a reject number of 8 is a borderline fail. 30 against 8 is a systemic problem.
• Whether the defects repeat — twenty units with the same wrong stitching is one process error. Twenty different defects is a factory quality-control breakdown, which is worse.
• Whether the sampling was sound — random draw from sealed export cartons, a defect log tallied against the AQL, real photos. If the report cannot show that, question the result before acting on it. Here is how to pressure-test a report.

Ten minutes with the defect log tells you whether you are managing a hiccup or a crisis — and everything below depends on which one it is.

The first 24 hours: three rules

1. Do not release the shipment

Tell your supplier, in writing, that the goods must not ship until you have reviewed the findings and agreed on next steps. Most suppliers will not ship a failed lot without authorization — but "most" is not "all," and once containers are loaded your leverage drops to near zero. If your payment terms have a balance due against shipment, do not release that either.

2. Do not accept a vague fix

The reflex response from many factories is "we will rework everything, don't worry." Rework without a defined scope, method and completion date is not a plan — it is a delay with extra steps. Wait until you have done the analysis below before agreeing to anything.

3. Get the full evidence while it is fresh

Ask your inspection company for all defect photos, the completed defect log, and the sampling details if they are not already in the report. If anything is unclear, query it now — inspectors' memories of a specific booking fade fast, and you may need this evidence for a negotiation later.

Your four options — and when each one makes sense

Every failed inspection resolves in one of four ways. The right choice depends on the defect class, the margin at stake, and your deadline.

Option 1: Rework and re-inspect (the default)

The factory repairs or replaces the defective portion, and a re-inspection verifies the result before shipping. This is the standard resolution for most major-defect fails.

Make the rework agreement specific, in writing: what will be reworked (which defects, which cartons — the whole lot, not just the inspected sample), how (the repair method — some "fixes" create new defects), by when, and how it will be verified (a full re-inspection, not the factory's word). A realistic rework cycle on a typical consumer-goods order is 3–10 days plus re-inspection; build that into your customer communication now rather than promising the original date and missing it.

Option 2: 100% sorting

Instead of reworking, every unit is checked and the defective ones are pulled. This fits when defects are easy to spot unit-by-unit and concentrated — for example, a scratch or misprint affecting part of production — and it is usually faster than rework. Many inspection companies offer sorting as a paid service with their own staff, which removes the fox-guarding-henhouse problem of a factory sorting its own output.

Option 3: Accept with a concession

If the fail is minor-class only and borderline — cosmetic issues your end customer is unlikely to return the product over — accepting the lot in exchange for a documented discount can be rational, especially against a hard deadline. Two cautions: get the concession in writing before releasing anything, and understand that every concession teaches your supplier your real quality bar. Do it twice and your AQL becomes decoration. Never accept a concession on a critical fail.

Option 4: Reject the lot

For critical-class failures (safety, legal compliance), systemic quality collapse, or a supplier who will not engage with corrective action — refuse the goods. What happens next depends on your contract and payment terms: this is where a purchase order that specifies inspection pass as a condition of payment earns its keep. Rejection is rare and expensive for everyone, which is precisely why the credible option of it keeps the other three honest.

Demand corrective action — properly

Fixing this lot is half the job. The other half is making sure the same defect is not waiting in your next PO.

Send your supplier a short, structured corrective-action request (CAPA). It does not need to be a 20-page quality document. One page:

1. The defect(s), quoted from the report with photos — not "quality problems," but "loose stitching at side seam, 14 units of 200, exceeds Re = 8."
2. A root-cause question, not an accusation: ask what in the process produced this — material, machine setting, operator step, missing internal check.
3. The fix and the prevention: what changes for the reworked lot, and what changes permanently.
4. Verification: the specific checkpoints you will add to the next inspection to confirm the fix held.

Two things happen when you send this. Suppliers who take quality seriously answer it — and their answer tells you a lot. And you stop sounding like a one-off buyer who can be waited out, and start sounding like a customer with a quality function. Factories triage their effort accordingly.

The re-inspection: who pays, and what changes

Who pays: the widely used convention — and a clause worth writing into your PO terms — is that the supplier bears the cost of re-inspection after a fail, since their goods caused it. Agree on this before the first inspection is ever booked, not during the argument afterward. In practice some buyers split or absorb it to preserve a relationship; that is a commercial call, but the default ask should be supplier-pays.

What changes: do not just run the identical inspection and hope. Two adjustments are standard practice:

• Target the failure. Brief the re-inspection explicitly on the failed defects and the rework agreement, so the inspector verifies the fix — not just the general condition.
• Consider tightening. ISO 2859-1's switching rules exist for exactly this situation: repeated failures move a supplier from normal to tightened inspection, where the same sample size carries a stricter accept number. For a one-off re-inspection after a single fail, many buyers simply re-run at normal severity with a targeted brief; if this supplier has now failed twice in five lots, tightened inspection is the by-the-book response. See how the sampling math works.

And use a fresh, complete checklist — if the original inspection missed checkpoints that should have caught issues earlier, fix the brief too. Our free product-specific inspection templates cover the checkpoints by category.

When to walk away from the supplier

One failed inspection is information, not a verdict — good factories fail inspections sometimes, and a supplier who responds fast with a serious corrective action often comes out more trustworthy than before.

Patterns are the verdict. Consider exiting when you see: repeated fails on the same defect across POs (they cannot or will not fix their process), critical failures involving safety or compliance, evidence of hidden or manipulated quality (pre-picked samples, blocked inspection access, pressure to skip inspection), or corrective-action responses that are excuses rather than process changes. And if you are sourcing from a country where you have one supplier and no bench, that is the deeper vulnerability — a China+1 strategy is the structural fix.

The checklist

1. Read the defect log — which class failed, by how much, repeating or scattered.
2. Freeze shipment and payment in writing.
3. Collect full evidence from your inspection company.
4. Choose: rework / sort / concession / reject — based on class and margin, not the calendar.
5. Put the rework agreement in writing: scope, method, date, verification.
6. Send a one-page corrective-action request.
7. Book the re-inspection — supplier pays, targeted brief, tightened if it is a repeat.
8. Log the outcome against this supplier for your next sourcing decision.

A failed inspection is not the failure — shipping the defects would have been. Handle the next few days deliberately: freeze, analyze, choose, document, verify. Do that once, and you will have a supplier who takes your quality bar seriously on every order that follows.

Frequently asked questions