Reading inspection reports without guesswork: part 11

Key takeawaysReports sort defects into critical, major, and minor because each class carries a different acceptance threshold, and mixing them up is the most common reading error^[1]. · The verdict compares sample findings against the agreed acceptance numbers, so the first page to check is always the sampling basis, not the photo gallery. · Acceptance testing logic treats the finished consignment as the unit under test: specified requirements are checked before any release decision, the same discipline factory acceptance testing applies to equipment handover^[2]. · Photos prove what words describe, so each finding worth acting on should trace to an image, a location in the consignment, and a stated requirement it violates. · Remarks and limitations define what the verdict does not cover, and skipping them is how buyers approve goods the inspector never actually saw.

Most buyers open an inspection report at the photos, scroll to the verdict, and then argue with the supplier about the middle. That order wastes the document. The middle, defect classification, counts, and the comparison against acceptance levels, is the reasoning; the verdict is only its last line. Read in the right order, even a failed report tells you exactly what to fix and whether a re-check is proportionate.

The classification behind every report comes from acceptance sampling practice, which defines how sample results map to lot decisions through statistically grounded acceptance and rejection numbers^[1]. A report that cites AQL 1.0 for critical defects and AQL 2.5 for majors is applying that framework to your cartons. Knowing what those numbers mean before the goods ship is the difference between negotiating from the report and negotiating around it.

This article walks through the three defect classes, the anatomy of a findings page, a decision table that maps patterns to next steps, and a short reading routine you can apply to any report. It assumes a standard pre-shipment inspection with an AQL basis already agreed. Later parts of this series cover compliance sections and laboratory references; here the focus is purely on workmanship and functional findings.

Critical, major, minor: what the three words actually decide

A critical defect is one that makes the product unsafe or unfit for its intended purpose: a live part accessible to the finger, a child seat buckle that releases under load, a food-contact coating that flakes into contents. Most acceptance plans set critical acceptance at zero, which means a single confirmed critical in the sample fails the consignment regardless of everything else. When you see the word critical in a report, stop reading for quantity and start reading for containment: which cartons, which production window, and whether the goods can be segregated.

A major defect impairs function, performance, or saleability without creating a safety hazard: a zipper that jams, a motor that stalls under rated load, a printed shade far outside the approved tolerance. Majors are counted against the agreed AQL, typically 1.5 or 2.5 depending on the product and the buyer's risk appetite. The number matters less than its trend across shipments, because a rising major rate is the earliest signal that a process, not a batch, is drifting.

A minor defect is a cosmetic or workmanship blemish that neither harms function nor blocks sale: an untrimmed thread, a faint molding mark in a hidden area, a slightly uneven print on an inner flap. Minors usually carry the loosest level, often 4.0, and they are where buyer-supplier arguments concentrate. The way out is a sealed reference sample plus a defect catalogue with photos, so minor versus acceptable is decided once at approval stage rather than argued on every shipment.

How to read the findings pages

Findings pages follow a repeating anatomy: requirement, observation, evidence, and classification. Each entry should name the requirement that was checked, describe what was observed, point to the photo or measurement that proves it, and state the class assigned. When any of the four is missing, the entry is a note rather than a finding, and you should treat it as a question to the inspector rather than a fact about the goods.

Match every count to its basis

Beside each defect count the report states the sample size it came from and the acceptance number for that class. Twelve majors mean nothing until you know whether the sample was eighty pieces or three hundred, and whether the acceptance number was ten or fourteen. Check the sampling table first; it is usually printed near the front and it governs everything after it^[1].

Watch for split results across strata, where one SKU or colour passes and another fails within the same consignment. A single blended verdict can hide a failing stratum, so ask for per-stratum counts whenever the order mixes models. If the report does not break them out, request the working notes before deciding.

Treat photos as cross-examination

A useful photo carries a caption with the carton number, the defect, and a scale reference such as a ruler. Flip between the finding text and its image and ask whether the image alone would convince a stranger: if it would not, the finding is weakly evidenced. Strong reports also include at least one overall view per checked carton, so selective photography cannot create a false impression of the lot.

Compare defect photos against the approved reference sample rather than against perfection. Production goods always differ from golden samples in trivial ways, and the report records deviations from agreed tolerances, not from an ideal. Where the tolerance itself was never written down, the finding exposes a specification gap to close before the next order, not just a factory error.

From finding pattern to proportionate response

Patterns like these are easier to compare when you know what a complete report looks like in advance. Buyers who want to check current report formats and service scope descriptions can find them at www.utsinspection.com before briefing their supplier on the next round.

A reader routine that fits on one page

1. Confirm the basis.Read the sampling standard, sample size, and acceptance numbers first, and verify they match what was booked. A verdict built on the wrong basis is a verdict about the wrong question.
2. Classify before counting.Walk the findings and check that each defect sits in the right class; a misclassified critical hides inside the major count more often than buyers expect. Reclassification changes verdicts.
3. Trace each finding to evidence.Every actionable finding should lead to a photo with a carton number and a named requirement. Findings without that trail go back to the inspector as questions.
4. Read the remarks and limits.Access restrictions, untested functions, and excluded SKUs live here, and they bound what the verdict actually covers. Factory acceptance practice is explicit on this point: a test record is only valid for the requirements and conditions it actually witnessed^[2].
5. Decide in writing.Record the disposition, release, conditional release, sort and re-check, or reject, with the report reference and the evidence relied on. A verdict without a written disposition decays into a dispute within a week.

Frequently asked questions