The template
Plan number: · Date raised: · Raised from (audit, NCR, incident, complaint, board measure): · Area / process: · Plan owner: · Target close date:
| Section | Field | Entry |
|---|---|---|
| 1. Problem | What happened, where, when, how often · Requirement not met (procedure, specification, standard clause, target) · Evidence (photos, data, records) · Source reference (NCR or audit finding number) | |
| 2. Containment | Immediate action already taken · Who, when · Is anything else at risk (other lines, shifts, batches, sites)? | |
| 3. Root cause | Method (5 whys, fishbone, data) · Root cause statement · Why it was not detected earlier | |
| 4. Corrective actions | One row per action, in the table below | |
| 5. Verification | Measure of effectiveness · Check date and period · Result · Verified by | |
| 6. Closure | Standard, procedure or checklist updated · Shared with (other teams or sites) · Closed by, date |
Action table
| # | Action (addresses which cause?) | Type (correct / prevent / standardise) | Hierarchy-of-control level (safety) | Owner | Due | Status | Evidence |
|---|---|---|---|---|---|---|---|
| 1 | |||||||
| 2 | |||||||
| 3 |
What is the difference between a corrective action plan and an NCR?
A non-conformance report records one failure to meet a requirement and what was done with the affected items. A corrective action plan is the part that removes the cause, and one plan can cover several NCRs, an audit finding or an incident. Many sites raise the NCR first and open a corrective action plan when the root cause needs more than a quick fix, or when the same NCR keeps returning. The NCR template on this site has the corrective action as section 6; use this plan when that section needs its own owner, timeline and review.
A filled example
| Field | Entry |
|---|---|
| Problem | CAP-0093 · Internal audit, 4 Sep · Forklift pre-start checks missing on 9 of 30 days sampled across two shifts · Requirement: site procedure SP-12, pre-start before first use each shift |
| Containment | Supervisor checked every forklift on both shifts the same day; one unit tagged out for a faulty horn |
| Root cause | Paper checklists kept in the office, 150 m from the forklift bay; night shift starts before the office opens; nobody reviews missing checks |
| Action 1 | Move pre-start to a phone checklist at the bay (shift supervisor, 11 Sep) |
| Action 2 | Missing pre-start shown on the Tier 1 board each morning (team leader, 11 Sep) |
| Action 3 | Same review at the two other warehouses (operations manager, 30 Sep) |
| Verification | 30 consecutive days with 100 per cent pre-start completion, checked 15 Oct; result met |
| Closure | SP-12 updated to rev 5; shared at Tier 3; closed by the site manager, 17 Oct |
How do you write a good corrective action?
- Tie each action to a root cause you listed. An action that matches no cause is a guess.
- Write it as a change to the process, not a reminder. "Retrain operators" on its own rarely stops a repeat; changing the checklist, the layout or the equipment does.
- For safety, aim as high on the hierarchy of control as is reasonably practicable and record the level.
- Give one person ownership and a real date. "Maintenance team, ASAP" is not an owner or a date.
- Decide how you will know it worked before you start, so verification is a check, not a debate.
Common mistakes
- Closing the plan when the actions are ticked off, before anyone checks the result.
- Root cause written as "human error" or "operator did not follow procedure".
- Containment recorded as the corrective action.
- Plans kept in a spreadsheet nobody reviews, so overdue actions sit for months.
- The same problem raised again next quarter under a new number.
Corrective action plans inside daily management
Open plans and their actions belong on the tier board, not in a separate register. The Tier 1 huddle sees actions due that day; the Tier 2 meeting reviews overdue actions and plans waiting for verification; repeat causes go up to Tier 3 as improvement projects or A3s. When the verification date arrives it is a dated action like any other, which is what stops plans being closed on paper only.
Step by step
- Describe the problem. Record what happened against the specific requirement, with data, photos and the source reference.
- Contain. Record the immediate action already taken and check whether other lines, shifts or sites are at risk.
- Find the root cause. Use 5 whys or a fishbone with the people who run the process, and ask why it was not detected.
- Plan the actions. One or more actions per root cause, each with a single owner, a due date and, for safety, the hierarchy-of-control level.
- Verify effectiveness. Check the agreed measure over the agreed period; if it has not held, go back to the root cause.
- Standardise and close. Update the procedure or checklist, share the lesson with other teams, then close the plan.
A worked example
A food manufacturer had 46 open corrective action plans from audits and customer complaints, tracked in a spreadsheet the quality team updated monthly. Half were overdue and a third had been closed without any check. The site moved every open action onto the area tier boards with a named owner, added a verification date to each plan, and put overdue plans on the Tier 2 agenda every Tuesday. Within three months the open list was down to 12, and two plans that had been closed on paper were reopened when their verification checks failed.