8D (Eight Disciplines) is what a root cause analysis turns into when a customer or a supplier is on the other end of it. Ford originally developed the method to give supplier quality issues a consistent, auditable structure, and it's since become the default format automotive, aerospace, and industrial suppliers use whenever a formal corrective action report is required.
If you've read a basic root cause analysis process before, most of 8D will look familiar: define the problem, find the root cause, fix it. The difference is what 8D adds on either side of that core: a containment step before the root cause is even known, and a prevention and documentation step after the fix is verified. Those additions exist specifically because 8D is built for situations where a defective part has already left the plant, or a customer is waiting on a written response.
This post covers all eight disciplines, a worked example from a supplier corrective action request, when 8D is the right level of process versus overkill, and where the workflow typically breaks down without automation.
Not every problem needs 8D. A machine that jams twice a shift is a good candidate for 5 Whys or a fishbone diagram run internally by the team that owns the line.
8D is the right tool when:
If none of those apply, a lighter RCA process gets you to the same root cause with less overhead. 8D's extra structure is the point when there's a customer or auditor who needs to see it.
Assign a cross-functional team with the authority and knowledge to solve the problem, not just the person who happened to notice it. For a supplier-facing issue, this typically includes quality, the process or design engineer responsible for the affected part, and someone from production.
State the problem in specific, measurable terms: what failed, on which part number, at what rate, discovered when and by whom. "Bracket cracking at the mounting hole, 3.2% of lot #4471, discovered during customer incoming inspection on [date]" is a usable problem statement. "Some brackets are cracking" is not.
Before the root cause is known, stop more defective product from reaching the customer or continuing through the process. This might mean 100% sorting of existing inventory, a hold on shipments from the affected lot, or an added inspection step at the next operation. The containment action is temporary by definition and gets removed once a permanent fix is verified.
This is the step most commonly skipped when teams try to shortcut 8D by jumping straight to root cause analysis. Skipping containment means more defective parts keep shipping while the investigation is still in progress.
This is where the core RCA work happens: 5 Whys or a fishbone diagram applied to find why the defect occurred. 8D asks for two root causes, not one: the occurrence root cause (why the defect happened) and the escape point (why it wasn't caught before it shipped, i.e., which inspection step should have caught it and didn't). Both need a permanent fix; fixing only the occurrence cause without fixing the detection gap means a different defect can still escape the same way in the future.
Select a fix that addresses the verified root cause, then confirm it actually works before rolling it out, typically through a pilot run or a trial period. This is distinct from D3's containment: containment stops the bleeding immediately, corrective action is the permanent fix that lets containment be removed.
Roll out the corrective action to full production and confirm the defect rate has actually dropped using real production data, not just confidence that the fix should work. Once validated, the interim containment action from D3 is removed.
Update the systems that let the defect happen in the first place: control plans, FMEAs (Failure Mode and Effects Analysis), work instructions, or the standard parts list, so the same failure mode can't recur on this part or on similar parts elsewhere in the plant. This is the step that turns one fix into an actual process improvement instead of a one-time save.
Close out the report, document the full timeline and outcome, and acknowledge the team's work. For supplier-facing 8Ds, this is also the point where the formal report gets submitted back to the customer.
D1 - Team: Quality manager, process engineer, production supervisor for the affected line.
D2 - Problem: Customer incoming inspection found cracking at the mounting hole on 3.2% of brackets in lot #4471 (140 of 4,375 units). Defect not present in the customer's inspection of the three prior lots.
D3 - Containment: Remaining inventory from lot #4471 (2,890 units) placed on hold. 100% visual inspection added at final pack for the next two production lots while the investigation is underway.
D4 - Root cause: Occurrence root cause traced (via fishbone, since multiple factors were plausible) to a stamping die showing early wear at the mounting-hole punch, increasing burr formation that led to stress cracking under load. Escape point: the in-process visual check specified in the work instruction does not include the mounting hole area, only the outer profile, so the defect wasn't caught until customer incoming inspection.
D5 - Corrective action selected: Die maintenance/resharpening on the mounting-hole punch station, verified on a 200-unit pilot run showing 0% cracking. Work instruction updated to add the mounting hole to the in-process visual check.
D6 - Validation: Full production run of 1,500 units following die maintenance and the updated work instruction shows 0 cracked units. Containment (100% inspection) removed after this validation.
D7 - Prevention: Die maintenance interval updated in the preventive maintenance schedule to catch wear earlier on this punch station. Same mounting-hole visual check added to the work instructions for two other parts run on the same die family.
D8 - Close: 8D report submitted to the customer with full timeline; team acknowledged at the next quality meeting.
Note the two root causes: the die wear explains why the defect occurred, and the incomplete visual check explains why it wasn't caught before shipping. A team that only fixed the die would have left the escape point open for the next unrelated defect.
The eight disciplines are a documentation-heavy process by design, which is exactly why teams tend to let containment actions, corrective action due dates, and validation checks slip once the initial report is filed.
An n8n workflow can carry the parts of 8D that depend on follow-through rather than analysis:
This sits on top of whatever you're already using to log nonconformances and defects, whether that's a quality management system, a spreadsheet, or your ERP's quality module. See Quality Control Automation on the Shop Floor for the broader pattern of connecting inspection data to an automatic response, and DPMO Formula: How to Calculate Defects Per Million Opportunities and Manufacturing KPIs: The Metrics Operations Teams Actually Track for how 8D outcomes should roll up into the metrics you're already tracking.
The Flow Kaizen guide covers how to build containment tracking, corrective action escalation, and report assembly workflows like the ones above using n8n.