8D vs. CAPA: When to Use Each

8D and CAPA are not competitors; they operate at different levels. CAPA (corrective and preventive action) is the quality-system process that governs how an organization eliminates causes of nonconformities, with owners, due dates, and effectiveness reviews. 8D (the Eight Disciplines) is a specific team-based method for running one such investigation, D0 through D8. In practice you run an 8D inside a CAPA record, not instead of it.

Plants confuse the two because they cover so much of the same ground: both find root cause, both verify a fix, both aim to stop recurrence. But one is a management system and the other is a technique, and mistaking the technique for the system, or the system for the technique, is how organizations end up with either 8D reports that nobody tracks to closure or CAPA records with no real problem-solving inside them. This is a comparison of two quality tools that work best together, not a contest between vendors.

What is CAPA?

CAPA is the closed-loop process a quality management system uses to remove the causes of problems so they do not recur. It is a requirement, not an option, across most quality standards: ISO 9001 covers it under nonconformity and corrective action (clause 10.2), ISO 13485 and FDA's device regulation carry explicit corrective and preventive action requirements, and IATF 16949 builds it into automotive quality management. A CAPA is fundamentally a governed record: a problem is logged, an owner is assigned, a root cause is found, actions are taken, effectiveness is verified after implementation, and the record is closed with evidence. The system's job is accountability, making sure every serious problem is investigated, fixed, and confirmed, and that none quietly fall through. Our full guide to CAPA walks the eight-step lifecycle in depth.

What is 8D?

8D is a structured, team-based problem-solving method that Ford Motor Company formalized in a 1987 manual called Team Oriented Problem Solving (TOPS), later expanded with a front-end planning step (D0) into Global 8D. It runs nine disciplines, D0 through D8: plan, build the team, describe the problem, contain it, find and verify the root cause, choose corrective actions, implement and validate them, prevent recurrence, and close by recognizing the team. It is the standard currency between automotive manufacturers and their suppliers: when an OEM receives a defect, it typically requires an 8D report back. Where CAPA is the filing cabinet and the calendar, 8D is the investigation that fills one folder. Our 8D problem solving guide covers each discipline.

Where do 8D and CAPA overlap?

They overlap almost entirely on substance, which is why they are so easy to confuse. Both are corrective-action loops that insist on the same core moves: describe the problem in measurable terms, protect the customer while you investigate, find and verify a real root cause rather than a symptom, act on that cause, confirm the action worked, and prevent the problem from coming back elsewhere. Both reject the shortcut of closing on "retrained the operator." Both end only when there is evidence the fix held. If you strip away the labels, an honest 8D and a well-run CAPA investigation contain the same thinking, because 8D was one of the templates the modern CAPA discipline was built from.

Where do they differ?

The differences are about scope and structure, not philosophy. CAPA is a governance system that can hold many kinds of investigation; 8D is one specific investigation format. A few distinctions matter in practice.

• System vs method. CAPA is the record and the accountability around it, owners, due dates, effectiveness checks, audit trail. 8D is the technique used to work the problem. You can run a CAPA with a simple 5 whys for a minor issue, or with a full 8D for a serious one.
• Team and containment are explicit in 8D. 8D mandates a cross-functional team (D1) and interim containment (D3) as named steps. A generic CAPA process requires neither by name, though good ones do both.
• The escape point. 8D explicitly investigates why the defect was not caught (the escape point) alongside why it was made. The generic CAPA framework does not name this, though it lives inside a thorough root-cause step.
• Verification timing. This is the subtle one. Classic 8D verifies chosen corrective actions before full implementation (D5), then validates with production data after rollout (D6). FDA's CAPA expectation is specifically an effectiveness check after implementation. The two are compatible, but a team using 8D under FDA rules has to make sure the post-implementation effectiveness verification is not skipped just because D5 already "verified" the action.
• Origin and reach. 8D comes from automotive and is the common supplier-to-OEM report format. CAPA is industry-agnostic and appears in medical devices, food, aerospace, and general manufacturing.

How do you run an 8D inside a CAPA?

The cleanest way to use both is to let CAPA govern and 8D investigate. A customer complaint or a serious internal nonconformity opens a CAPA record; the investigation assigned to that record is run as an 8D; and the CAPA stays open until the 8D's validation data and preventive actions are confirmed. Here is the sequence.

1. Trigger and log the CAPA. The complaint, audit finding, or nonconformity opens a CAPA record with an owner and a due date. This is the accountability layer that will outlast the investigation.
2. Decide the depth. Triage by risk. A minor, one-off issue may only need a quick 5 whys inside the CAPA. A recurring or customer-facing problem justifies a full 8D. Not every CAPA needs the full nine disciplines.
3. Run D0 through D3. Plan, build the team, describe the problem in measurable terms, and contain it. The containment action is recorded in the CAPA as a temporary measure, explicitly not the fix.
4. Run D4 through D6. Find and verify the occurrence and escape root causes, select corrective actions, implement them, and validate with production data. Feed each into the CAPA's action list.
5. Do the CAPA effectiveness check. After implementation, verify with dated evidence that the problem stopped and no new problem appeared. This satisfies both D6 validation and the CAPA's own post-implementation requirement, the one step teams most often let slip.
6. Run D7 and D8, then close. Prevent recurrence on sister lines and similar products, standardize the fix, recognize the team, and close the CAPA on evidence, not on the calendar.

When should you use each?

Use 8D when the problem is recurring, high-impact, or customer-facing, when the cause is genuinely unknown, and when a cross-functional team is needed, especially if a customer or OEM is expecting an 8D report. Use a lighter method inside CAPA, such as a 5 whys, for a one-off with an obvious cause. And remember that CAPA is always the governing layer in a regulated quality system: even the lightest investigation belongs in a tracked record. Matching the method to the problem keeps the system fast enough that people take it seriously.

What do the standards and origins say?

How does a connected floor help both?

Both 8D and CAPA depend on the same fuel: evidence captured as events happen, not reconstructed from memory a week later under customer pressure. When downtime reasons, quality checks, and process parameters are timestamped at the station, D2 (describe) and D4 (root cause) start from real data, and the CAPA effectiveness check runs against a live stream instead of a manual tally. Plants that moved from paper logs to real-time operational data, like the processor in our CLS case study argue about causes with evidence, which is exactly the condition both methods need to close honestly. That visibility also surfaces the repeating fault signatures that tell you a fix on one line should be spread across the plant, tying the investigation back to the cost of quality it exists to reduce. See how Harmony captures floor data without a rip-and-replace on our features overview.