Pass, fail, and hold are decisions, not decorations. Each verdict covers only its stated scope and demands specific next steps: passes release defined batches, fails trigger containment and diagnosed correction, holds freeze goods until named evidence arrives. Reading reports this way turns paperwork into control.
Key takeaways
- Read the verdict together with its scope: a pass covers only the inspected characteristics, batches, and methods stated in the report.
- Sampling verdicts follow acceptance logic: the inspected sample represents production made under the same conditions [1].
- Acceptance decisions need agreed criteria before the visit, the same discipline factory acceptance practice demands [2].
- A fail triggers containment first: segregate affected batches, then require diagnosed correction with deadlines.
- A hold is a pending verdict with a defined release condition; treat held goods as unshippable until it is met.
Every during production inspection ends in a report, and every report ends in a verdict: pass, fail, or hold. Yet buyers routinely misread all three. Passes get treated as blanket approvals, fails trigger emotional standoffs instead of structured correction, and holds quietly dissolve into shipments nobody authorized.
Misreading is expensive. A pass misunderstood as total approval lets unchecked risks sail; a fail mishandled as an insult destroys the correction cooperation the buyer needs; a hold left undefined becomes a shipment with a paper trail of warnings nobody acted on.
This guide teaches the decision grammar of production reports: what each verdict means, what it does not mean, and the exact next steps each one demands from the buyer and the factory.
Anatomy of a usable report
A usable report opens with identification: order, product, factory, date, inspector, and the production window observed. Without this header, findings float without attachment to any lot. Check first that the report describes your order and not a neighboring one; mistaken identity happens more often than buyers expect.
The body must link every finding to quantities and batches: defect description, photos with scale, counts found, batch numbers affected, and production dates. Findings without quantities cannot drive sorting or rework decisions. Findings without batch links cannot be contained. Insist on both before accepting the report as complete.
The verdict section states pass, fail, or hold per inspected scope, lists exclusions, and names pending items with owners and dates. Exclusions matter as much as verdicts: processes not running, materials not yet delivered, tests awaiting lab results. A report that hides its own limits invites decisions its evidence cannot support.
What each verdict means and demands
| Verdict | Meaning | Buyer next step |
|---|---|---|
| Pass | Inspected scope met the agreed criteria | Release the covered batches; schedule follow-up for excluded scope |
| Conditional pass | Criteria met except named open items | Release only after open items close with recorded evidence |
| Fail | Criteria not met in the inspected scope | Contain affected batches; demand diagnosed correction with deadline |
| Hold | Verdict pending defined evidence | Freeze affected goods; set the release condition, owner, and date |
| Scope-limited pass | Only listed checkpoints verified | Treat unverified areas as unknown, not as approved |
Handling fail without breaking the relationship
Separate the commercial relationship from the technical verdict. The fail describes the goods against agreed criteria; it accuses nobody of bad faith. State it that way in writing: the sample showed these deviations from these criteria, therefore these batches are held pending these corrections. Factories cooperate with precise technical claims and resist vague quality complaints.
Demand diagnosis, not just rework. Re-sorting defective units without finding the cause returns the same defect in the next batch. Require the factory to name the cause, change the input or process behind it, and present evidence the change took effect. The correction record should read like an engineering note, not an apology letter.
Keep the door open for the factory valid objections. Agreed criteria sometimes contain errors, and inspectors sometimes misread drawings. A joint re-check of the disputed sample with the criteria on the table resolves honest disagreements fast. Conceding a genuine error strengthens authority over the findings that stand. Acceptance practice in industrial supply treats this adversarial collaboration as normal procedure [2]: verify, dispute, resolve, re-verify.
Making hold verdicts stick
A hold must name four things: which batches and quantities are held, where they are stored, what evidence releases them, and by when. Holds missing any of these leak within days as goods drift toward packing. Get the factory written acknowledgment of all four on the visit day; memory of verbal holds fades exactly when it matters.
Assign one owner for the hold on the buyer side. Shared responsibility for held goods means nobody checks them. The owner tracks the pending evidence, chases the laboratory or the missing information, and authorizes or refuses release in writing. When the evidence arrives, the release note should reference it by number and date.
The statistical basis deserves respect here: the hold exists because the sample verdict cannot yet be extended to the batch. Acceptance sampling theory treats the sample as representative of production under the same conditions [1], and a hold simply says the representation is incomplete until pending checks close. Releasing held goods early discards that logic and gambles the shipment on hope.
Closing the loop with verified correction
Scope re-verification around the correction, not around the calendar. Fresh samples from the corrected configuration, covering the failed characteristics and adjacent ones the remedy may have touched, with batch linkage recorded. Re-checking unrelated stock on an unrelated day proves the correction how? It does not. Write the re-check scope before booking it.
Require the factory correction record before the re-visit: cause identified, remedy applied, quantities reworked, process change documented. Re-verification without this record inspects effects without knowing causes, and causes unknown tend to recur. The record also becomes the training material that prevents the next order from repeating the failure.
File the closed loop per lot: original finding, correction record, re-verification verdict, with dates and quantities linked. This file answers the questions disputes always raise: what was wrong, what was done, and what proved it fixed. The service behind a UTS in-production quality check earns its keep in this loop, where independent verification turns factory claims into buyer-grade evidence.
Sources
- Wikipedia: Acceptance sampling —— Overview of acceptance sampling including AQL concepts used in product inspection.
- Wikipedia: Factory acceptance test —— Overview of factory acceptance testing performed before shipment or installation.
Frequently asked questions
What does a pass verdict actually guarantee?
Pass means the inspected portion met the agreed criteria within the stated scope; it does not bless unchecked processes or untested characteristics. Read the scope limits before celebrating. Conditional passes with open verifications still need follow-up before shipment.
What is the correct response to a fail verdict?
Hold the affected batches, prevent them from entering packing, and require the factory to identify the cause and present corrected units. Record quantities, batch numbers, and deadlines in writing. Release the hold only against conforming re-verification, never against verbal promises.
When is a hold verdict used instead of pass or fail?
A hold means the verdict is pending: usually laboratory results, missing information, or an incomplete check. Treat held goods exactly like failed goods until the pending evidence arrives: segregated, unpacked, unshipped. A hold is a pause with a defined release condition, not a soft pass.
How should re-inspection after correction be scoped?
Select fresh units from the corrected configuration with recorded batch linkage, and re-check the failed characteristics plus any adjacent ones the correction may have touched. Confirm the quantities covered and record the new verdict. Re-inspection of unrelated characteristics on unrelated batches proves nothing about the correction.