Quality Issue Severity Levels: A Triage System That Actually Works

Without a severity framework, every quality issue gets treated the same way: as a crisis or as background noise, depending on who reported it and how loudly. That's how minor defects get crisis treatment (burning out the team) while genuinely critical issues get under-reacted to (because the team is exhausted from chasing the minor ones). A simple triage system fixes this in a week.

Here's the four-level structure that works, with clear criteria for each level and a defined response.

Level 1: Critical (Safety or Recall)

Defects that could cause physical harm to an end user, regulatory non-compliance, or require a product recall. Examples: contamination in food/pharma, electrical safety failure, dimensional defects in safety-critical components. Response: stop production immediately, quarantine all related stock, notify the owner and the customer within 2 hours, full investigation within 24 hours. This level should fire less than once a year.

Level 2: Major (Customer-Affecting, Single Batch)

Defects that will be noticed by the customer and reasonably reflect on the brand, but don't pose safety or compliance risks. Examples: visible cosmetic defects on a finished product, wrong specification on a critical dimension, mislabelling. Response: stop dispatch of the affected batch, investigate within 24 hours, communicate proactively to the customer before they discover it, decide on rework vs. credit within 48 hours.

Level 3: Minor (Internal, Process Drift)

Defects caught internally that don't reach the customer — process variation, a single bad unit in a sample, a deviation from spec that's still within tolerance. Response: log in the QC system, root cause within a week, fix in the spec sheet or process. No external communication. No production stop.

Level 4: Observation (Watch-List)

Not a defect, but a signal — a trend in inspection data, an operator's verbal report of an unusual noise, a near-miss. Response: log it, monitor, escalate only if a pattern develops. The point of capturing observations is to detect Level 3 issues before they become Level 2.

Why the Framework Matters

The single most expensive failure mode in SME quality systems is mis-classification. Treating Level 3 issues as Level 1 burns out the team. Treating Level 2 issues as Level 3 loses customers. A clear severity framework, written down and trained, is what aligns the team's response intensity to the actual stakes.

Who Decides the Level

The person who detects the issue proposes a level. A second person (the shift supervisor or QC lead) confirms or upgrades. Downgrades require manager approval. This two-person rule prevents both panic escalation and quiet underplay.

Review the System Quarterly

Look at your last quarter's classifications. Were they consistent? Did anything that should have been Level 2 get marked Level 3? Did anything fire Level 1 that shouldn't have? Recalibrate the criteria based on what you actually saw. The framework only stays useful if it tracks reality.