NCR management — handling material and line rejections properly

What an NCR actually is

A non-conformance report — an NCR — is the record of any product or process that fails to meet a requirement. That requirement might be a drawing dimension, a specification limit, a material property, a visual standard or a customer-specific rule. The NCR is not the defect; it is the controlled record of the defect: what was non-conforming, how much, where it was found, what defect it carried, what you did about it, and whether it is closed.

For an IATF 16949 or ISO 9001 plant, control of non-conforming output is a core clause. The standard does not just ask that you find defects — it asks that you identify and control non-conforming product so it cannot be used or shipped by mistake, that you take appropriate action, and that you keep records of the non-conformity and the action taken. An NCR process is how you satisfy all three. Without it, a rejection is a note in a book that no one analyses and an auditor cannot follow.

The three kinds of rejection

Not all rejections are the same, and conflating them is the first mistake. A robust quality system distinguishes at least three, because each carries different data and drives different action.

The distinction matters because it is the only way to answer “whose problem is this?” A material rejection points at a supplier; a line rejection points at a process or work centre; a child-part rejection points at a specific component inside an assembly. In Fast Quality these are separate, purpose-built entries — material rejection at receiving, line rejection and rejection against the work order in production, and child-part rejection for assemblies — so the data lands where the action must be taken.

Defect codes — the common language

The single most valuable discipline in NCR management is a controlled defect-code catalogue. When every rejection is tagged from the same list of defect codes — “bore oversize”, “porosity”, “burr”, “wrong hardness”, “short-fill” — the defect becomes a common language that links receiving, the line, the 8D and the FMEA. Free-text rejection reasons, by contrast, are useless in aggregate: “bad”, “NG”, “not ok” and “rejected by QC” can never be Pareto-charted.

A good defect catalogue does three jobs at once. It makes a rejection Pareto possible, so you can rank defects by frequency or cost and attack the biggest first. It links a defect on the line back to the failure mode in the FMEA that predicted it and the special characteristic on the control plan meant to catch it. And it standardises the language of the 8D. In Fast Quality, the defect masters — the defect-code catalogue and defect-vs-work-centre mapping — are the shared vocabulary tying rejection to 8D to FMEA.

The NCR lifecycle, start to close

A non-conformance that is merely logged and forgotten is worse than none — it creates the illusion of control without the substance. A disciplined NCR moves through five stages, each a real action so the whole flow is auditable.

The stage plants most often skip is containment. It is tempting to record a rejection and move on, but if the non-conforming stock is not physically segregated, the odds that some of it reaches production or a customer are uncomfortably high — and an auditor who finds rejected material sitting next to good stock will write you up regardless of how tidy your paperwork is.

When a minor NCR becomes a deviation and a major one becomes an 8D

Not every non-conformance deserves the same weight of response. Applying an 8D to a single cosmetic blemish wastes the tool and buries the team in paperwork; treating a safety-characteristic failure as a routine reject is negligent. The judgement is in the triage, and it usually resolves into two paths.

The critical rule: a defect that keeps recurring is never minor, however small each instance looks. The third time the same defect code appears on the Pareto, it has earned an 8D — because the cost of a recurring defect is not one rejection, it is the sum of every rejection it will cause until someone finds the root cause. In Fast Quality, a minor NCR is dispositioned as accept-under-deviation through a controlled deviation-approval step, while a major or recurring one escalates into an 8D whose CAPA is driven through change management back into the FMEA and control plan.

Rejection data and supplier claims

An incoming rejection is not just a quality event — it is a commercial one. When a supplier ships you a defect, that defect costs money: the material itself, the sorting and containment labour, the delay, and sometimes the downstream scrap. A mature NCR process turns that cost into a supplier claim rather than absorbing it silently.

Because every material rejection is tagged with the supplier and a defect code, three things become possible that a spreadsheet cannot deliver. You can raise a supplier claim to recover the cost of the defect. You can maintain a real supplier scorecard — rejection PPM by supplier over time — that reflects what suppliers actually ship, not what their certificates claim. And you can decide, on evidence, when a recurring problem warrants an 8D rather than another return-and-replace. In Fast Quality, claim reporting runs off the material-rejection data, so the claim and the scorecard are by-products of the rejection you already recorded.

Common NCR mistakes

• Closing without root cause. An NCR closed with “re-inspected, found ok” or “operator counselled” is not closed — it is deferred. Recurring defects demand a real D4.
• No containment. Recording a rejection without physically segregating the stock is the fastest route to shipping a known defect and failing an audit.
• Deviation as a habit. When AD becomes the default for anything inconvenient to reject, the specification stops meaning anything.
• Free-text reasons. Without a controlled defect code, the rejection cannot be Pareto-charted and the whole dataset is analytically worthless.
• No feedback to the FMEA. A CAPA that fixes the line but never re-scores the FMEA leaves the risk analysis lying about a risk you just proved was real.

How Fast Quality manages non-conformance

Fast Quality Software, built by Improsys in Pune on the shared Fast Suite platform, implements the NCR loop as linked, purpose-built records rather than one generic form:

Because it runs on one shared platform, a customer complaint from Fast Complaint escalates natively into an 8D, and the loop back to the FMEA and control plan needs no middleware. For the full closed loop, see the pillar guide on what quality management software is.

Frequently asked questions