The 8D steps explained
The eight disciplines are ordered on purpose: each step depends on the quality of the one before it. A weak problem statement in D3 will corrupt every root cause candidate in D4. This page describes what actually happens in each step in a real investigation, and where each step commonly fails.
Key takeaways
- D1: pick the smallest team that can actually decide, not the largest that could attend.
- D2: contain based on suspect population, not on confirmed defective units.
- D3: describe the problem in Is / Is-Not terms with quantities and dates.
- D4: prove the root cause by turning the defect on and off — not just by explaining it.
- D5–D6: verify effectiveness with production data, not with lab tests alone.
- D7–D8: update PFMEA, control plan and lessons learned before closure.
D1 — Establish the team
Assemble a cross-functional team with process knowledge, product knowledge, customer knowledge and the authority to change things. A team of ten with no decision-makers will not close an 8D.
- Sponsor or champion (authority to allocate resources).
- Team leader (owns the 8D end-to-end).
- Subject-matter experts: process, product, quality, supplier.
- Customer contact (or single interface to the customer).
D2 — Contain the problem
The goal is to make sure no additional defective product reaches the customer while root-cause work is in progress. Containment scope must cover the whole suspect population, not just confirmed defective units.
- Quarantine suspect inventory at plant, distribution and customer sites.
- Add 100% sort/inspection at the escape point.
- Notify affected customers with an interim response.
- Track effectiveness of containment daily — not once at the end.
D3 — Describe the problem
D3 is where most 8Ds go wrong. Use Is / Is-Not to describe what the problem is and, critically, what it is not. Quantify.
- What: the exact defect signature (dimension, feature, mode).
- Where: which line, cavity, tool, shift, plant, region.
- When: first observed, last observed, frequency, pattern.
- How many: units affected out of total produced or shipped.
D4 — Identify the root cause
Identify the root cause of the defect (why it happened) and the escape cause (why controls did not catch it). Both must be proven, not asserted.
Proof means you can reproduce the defect by turning the suspected cause on, and prevent it by turning it off. If you cannot switch the defect on and off, you have a hypothesis — not a root cause.
D5–D6 — Corrective action and verification
D5 selects permanent corrective actions. D6 implements them in production and verifies that the defect does not recur. Verification uses production data over a defined observation window — not a one-off lab result.
D7–D8 — Prevent and close
D7 updates the systems that allowed the escape: PFMEA, control plan, work instructions, poka-yoke, supplier requirements, training. D8 closes the case, recognises the team and captures lessons learned so a similar problem does not require a new 8D.
Frequently asked questions
- What is the difference between root cause and escape cause?
- Root cause explains why the defect was produced. Escape cause explains why detection controls did not catch it. Every 8D needs both.
- How many people should be on an 8D team?
- Usually 4–7. Small enough to decide, cross-functional enough to cover process, product, quality and the customer interface.
- Can I run D2 and D3 in parallel?
- Yes — containment and problem description almost always run in parallel in the first 48 hours. What you cannot do is start D4 before D3 is quantified.
Related guides
Run this D-step inside a structured workspace
Qhubio is a structured 8D workspace: evidence-grounded phases, phase-gated approvals and audit-ready exports.