Managing Non-Compliances & Root Cause CAPA Submissions

An audit report is not the end of a compliance story. It is the beginning of the only part buyers remember: what the factory did next. Every finding in a SMETA, WRAP, BSCI or buyer code of conduct report comes with an expectation that the factory will respond with a corrective action plan, known as a CAPA. The CAPA is the document that converts criticism into improvement, and it is also the document most often done badly.

The pattern of weak CAPAs is remarkably consistent across factories. The action restates the finding instead of fixing it. Training is promised for every problem, whether or not knowledge was the cause. No owner is named, no deadline is set, and the evidence is described as something that will be provided later. Auditors see these patterns immediately, and they respond the only way they can: the finding stays open, the next audit starts with a credibility deficit, and repeat findings are graded more severely.

A strong CAPA is built on a simple discipline. Separate the immediate fix from the systemic fix. Prove the cause with a method, not a guess. Give every action a named owner, a calendar date and evidence that can be checked. Then verify that the fix actually works over time. This lesson teaches that discipline step by step, so that CAPA submissions become a factory strength rather than a recurring embarrassment.

Not all findings carry the same weight, and the first CAPA skill is reading the audit report accurately. Programmes use different grading words, so always confirm the grading scale of the audit in question rather than assuming it matches another programme. As a general working picture, the most serious findings involve immediate risk to workers, such as blocked emergency exits, child labour or forced labour indicators, or deliberate falsification of records. The next tier covers systemic failures, such as payroll that does not match attendance, missing safety equipment or uncontrolled overtime. The lowest tier covers isolated lapses and documentation gaps.

Grading matters because it sets the response order. Serious findings need immediate containment, often on the day of the audit itself, and senior management involvement. Systemic findings need a proper root cause analysis and a process change. Minor findings still need a CAPA, but they can be handled through routine correction and a short verification. A common factory error is to treat all findings as equal and work through them in report order, which leaves the most serious issues waiting while minor paperwork is polished.

Language matters too. Copy the finding wording exactly into the CAPA tracker rather than softening it. When a factory rewrites a finding to sound less serious, auditors notice, and trust falls. The CAPA should show the factory facing the finding squarely, with the auditor evidence reference attached, so that anyone reading the submission can trace it back to the original report.

The single most important distinction in CAPA writing is between three words that factories often blur together. A correction fixes the one instance the auditor found. If five workers were underpaid, the correction is paying those five workers the shortfall. A corrective action removes the cause so the problem does not happen again. In the same example, the corrective action might be linking biometric attendance directly to payroll so that the two records cannot diverge in the first place. A preventive action goes one step further and guards against the same failure arising in a related area, such as adding a pre-payroll reconciliation check to every department, not only the one audited.

Auditors test each layer separately. They ask whether the instance is fixed right now, whether the cause has been removed, and whether the fix has been verified over a sensible period. A CAPA that offers only a correction will be sent back with a request for the systemic action. A CAPA that offers only a grand systemic plan, with no immediate containment, will be questioned on what happens to the affected workers in the meantime. Every finding needs both layers, written as separate actions with separate evidence.

The classic failure mode is the universal training promise. Training is a legitimate corrective action when, and only when, lack of knowledge is the proven root cause. When the cause is a missing procedure, an unclear responsibility, a broken roster system or a supervisor under shipment pressure, training changes nothing. A credible CAPA matches the action to the cause that the analysis actually proved, and says so explicitly.

Root cause analysis is where most CAPAs fail, because it is tempting to stop at the first plausible explanation. Two methods, used together, prevent that. The first is the 5-Why chain: ask why the problem occurred, then why that reason occurred, and continue until the answer points to a process, a responsibility or a control that can actually be changed. The worked example in this lesson shows how an overtime finding traces back, through late fabric and a missing booking owner, to an order-start procedure with no deadline. That final answer is a root cause because fixing it changes the system. The earlier answers are symptoms, and fixing any of them alone would leave the problem free to return.

The second method is the fishbone review, which checks six families of causes: method, manpower, machine, material, measurement and environment. It is most useful for complex findings where a single 5-Why chain feels too narrow. A chemical storage finding, for example, might involve method (no segregation procedure), manpower (no trained storekeeper), material (unlabeled decanted containers) and environment (a store without bunding). The fishbone prevents the team from fixing one branch while the others keep producing the same finding.

Two rules govern both methods. First, never let the chain end at an individual worker. Blaming one person is not a root cause, because it leaves the supervision, the procedure and the system that allowed the failure completely untouched, and auditors treat worker-blaming as a failed analysis. Second, test the proposed root cause with a simple question: if this cause were fixed, would the problem stop recurring? If the honest answer is no, or only partly, keep digging.

A CAPA that survives verification has seven elements for every finding, and the absence of any one of them is a common reason for rejection. The finding is restated exactly, with its audit reference. Containment is described with a date, showing what was done immediately. The root cause is shown with its method, not merely asserted. The action is specific: which procedure changes, which system changes, which control is added. An owner is named as a person, not a department. A deadline is a calendar date, not a vague period. And the evidence is named precisely: which document, which log covering which period, which photograph, which record.

Evidence deserves special attention, because it is where factories most often promise more than they deliver. Strong evidence is dated, specific and produced by the normal operation of the fixed process. A revised procedure with approval signatures, three months of booking logs, overtime records covering the period after the fix, and training attendance sheets all count, because they show the fix living in daily work. Weak evidence is a statement that evidence will be provided later, a single photograph with no context, or a document created only for the auditor and never used again. Verification teams are experienced at telling the difference.

Deadlines also need honesty. If a deadline cannot be met, communicate before it passes, explain the reason, show the interim containment, and agree a revised date. Early, honest rescheduling is normally accepted. Silent slippage is recorded as poor CAPA management and damages the credibility of every other submission from the same factory.

Individual CAPAs fail without a system behind them. The most effective factories run one living CAPA tracker for the whole site, reviewed weekly by management, covering every audit source: SMETA, WRAP, BSCI, buyer inspections and internal audits together. The tracker shows each finding, its grade, its owner, its deadline, its evidence status and its verification result. When a new audit arrives, the tracker already shows which past findings are closed, which are overdue and which have repeated, and that history is itself strong evidence of a functioning management system.

Repeat findings deserve special treatment. A finding that reappears means the previous corrective action did not hold, and the correct response is to reopen the original CAPA, examine why the earlier action failed, escalate ownership to senior management, and add a stronger control with a longer verification period. Factories that quietly write a fresh CAPA for a repeat finding, as though the first one never existed, lose auditor confidence quickly. Factories that openly analyse the failure of their own previous fix, and change approach because of it, demonstrate exactly the management maturity that audits are designed to test.

Closure itself should be a decision, not a declaration. The factory should not mark a finding closed merely because the action was taken. Closure requires the evidence to have been checked and the fix to have held over a sensible period. Internal verification before the external verifier arrives is good practice: the compliance team re-checks the evidence pack, visits the affected area, and confirms the change is visible in daily work. That internal check catches weak evidence while there is still time to strengthen it.

Consider a complete cycle for a payroll finding. The audit reports that attendance registers and payroll records diverge for one section in one month. Containment happens immediately: the affected workers are paid any shortfall, and hand corrections to attendance sheets are suspended the same day. The root cause analysis finds that manual sheets were corrected after payroll was run, with no reconciliation step between the two records. The corrective action links biometric attendance to payroll and adds a monthly reconciliation check signed by HR and finance before payroll release. The preventive action extends the same reconciliation to all sections. Evidence includes the revised procedure, the biometric linkage record, three months of signed reconciliations and payslips showing zero divergence. Verification, after two clean months, confirms closure.

Notice what this cycle does not contain. There is no blame on the payroll clerk. There is no promise of training as a substitute for a system change. There is no vague deadline. Every claim is backed by a record that the normal operation of the factory now produces. That is the standard to apply to every finding, whether it concerns wages, safety, environment or working hours.

CAPA is where audits turn into improvement or into repeated failure. The discipline is not complicated, but it must be complete: face each finding exactly as written, contain the immediate instance, prove the root cause with a method, write actions that change the system, name an owner and a date for every action, and back every claim with checkable evidence. Then track all of it in one living register, review it weekly, verify closure properly, and treat repeat findings as system failures that demand a stronger response. Factories that follow this discipline close audits faster, face fewer findings at the next visit, and build the kind of management credibility that buyers reward with continued orders.