Corrective Action Plan Template — Free Download

Corrective Action Plan Template (D5-D6-D7)

Corrective actions fail for one reason more often than any other: nobody defined what "fixed" means before implementing the fix. This template gives you the three tables that make a corrective action auditable — a verification plan (D5), an implementation plan (D6) and an effectiveness monitoring plan (D7).

Use it whenever you close out an 8D report, respond to a SCAR, or manage a CAPA record under IATF 16949 or ISO 9001.

D5 — Verification Plan

Before full implementation, prove that the action actually addresses the root cause.

Rules for a good D5:

• Success criteria must be numeric — "no leaks after 1,000 h" beats "better sealing".
• State the test method so anyone can reproduce the verification.
• Choose a defensible sample size — enough units to detect the defect rate you are targeting.
• Record the measured result, not an opinion.

D6 — Implementation Plan

Once verification passes, roll the action out with owners, dates and risks.

Rules for a good D6:

• One owner per action — shared ownership is no ownership.
• Date every action and review overdue items weekly.
• Identify what the action touches — SOPs, control plans, FMEA, training, tooling, suppliers.
• Assess the risk of the change itself — a new material or process can introduce a new failure mode.

D7 — Effectiveness Monitoring

Verification proves the fix worked in a test. Monitoring proves it keeps working in production.

• 30-day check: ____
• 90-day check: ____
• 12-month check: ____
• FMEA updated: [ ] Yes [ ] No
• Control plan updated: [ ] Yes [ ] No
• Horizontal deployment reviewed: [ ] Yes [ ] No

How to Define Success Criteria

A corrective action is only verifiable if the success criterion is measurable.

If you cannot express the criterion as a number or a pass/fail test, it is not ready for D5.

Worked Example: Seal Leak Corrective Action

A pump housing seal leaks at end-of-line leak test. The team traces the root cause to a seal compound that hardens below the specified operating temperature. Here is how the three tables fill in.

D5 verification. The proposed action is a new seal compound. Success criterion: 0 leaks after 1,000 h at −20 °C to +90 °C cycling. Test method: pressure decay test at 1.5 bar, leak rate limit 0.5 cm³/min. Sample size: 30 units from an independent production lot, not the lot that failed. Result recorded as measured values per unit, with the worst-case unit called out explicitly.

D6 implementation. Actions: update the seal drawing and supplier specification (owner: Design Engineer, week 2); qualify the new compound with the supplier (owner: SQE, week 3); revise the incoming inspection plan (owner: QE Lead, week 4); update the SOP and retrain the line (owner: Production Supervisor, week 5). Each row carries a single name, a date and a risk note — the new compound is harder, so assembly force may rise and a new damage mode must be watched.

D7 monitoring. 30-day check: 100% leak test on all units, zero escapes. 90-day check: defect rate below 50 PPM over three consecutive lots. 12-month check: field return rate compared with the pre-fix baseline. FMEA and control plan updated, and the same seal family on other models reviewed for horizontal deployment.

Step-by-Step Checklist

1. Confirm the root cause is verified (D4) before writing any D5 row.
2. Write each success criterion as a number or a pass/fail test.
3. Name the test method, gauge and acceptance limit so the test is reproducible.
4. Select sample units independent of the failing lot.
5. Record measured results, including the worst case, not a summary opinion.
6. Assign exactly one owner and one due date per D6 action.
7. List every document the action touches: SOP, control plan, FMEA, training records, tooling, supplier PPAP.
8. Assess the risk the change itself introduces.
9. Define the 30-day, 90-day and 12-month monitoring checks with owners.
10. Update the FMEA and control plan and record the revision.
11. Review horizontal deployment across similar products and lines.
12. Close the 8D only when the D7 checks have passed.

Common Failure Modes and How to Detect Them

• Vague success criteria. Detect by asking a second engineer to run the test from the written criterion alone; if they cannot, rewrite it.
• Verification on the failed sample. Detect by checking lot and serial traceability of the D5 units against the D3 containment records.
• Orphaned actions. Detect by scanning the D6 table for blank owner or date cells — any blank is an open finding.
• Silent document drift. Detect by comparing the FMEA and control plan revision numbers before and after the action.
• Monitoring that never happens. Detect by auditing D7 check dates against the calendar; a check with no recorded result is not a check.
• Premature closure. Detect by confirming every D7 row has a measured result before the report status changes to closed.

How to Verify the Output

Before signing off, confirm that each D5 row has a numeric criterion, a method and a measured result; that each D6 row has one owner, one date and a risk note; and that each D7 row has a date, an owner and a recorded outcome. Cross-check that the FMEA and control plan revision numbers changed where the action required it, and that the monitoring sample is drawn from normal production, not from a hand-built trial batch.

Links to Neighbouring D-Steps

D5 depends on a verified D4 root cause — if the cause is still a guess, the verification plan tests the wrong thing. D6 depends on a passed D5; implementing before verification turns a trial into a production risk. D7 feeds back into D2 problem description and D4 analysis: if monitoring shows recurrence, the root cause was incomplete and the 8D reopens rather than closing. D7 also feeds the FMEA and control plan, which is how the lesson survives the next staff change.

Common Pitfalls

• Closing the 8D at D6 — without D7 monitoring, recurrence is invisible until the customer finds it.
• Verifying with the same sample that failed — use independent units.
• No FMEA or control plan update — the lessons do not survive staff turnover.
• Action lists without owners — every row needs a name and a date.

→ Skip manual planning — the free AI 8D Generator at report.8d.wiki drafts corrective actions, owners and verification plans for you.