Quality in Practice — Lesson 3: 8D report: complaint as a project, not a form

Quality in Practice · 5 min read · 08.09.2026

Welcome to the third lesson of the Quality in Practice series. In the first two, we showed that quality is created in the process, and how to reach the true root cause with the 5 Whys and Ishikawa. Today we bring both ideas together where the customer insists on them: in the complaint. We will walk through all eight steps so the complaint runs like a small project — with a team, deadlines, and verification. Not a form signed and forgotten.

When all that is left of 8D is a form

The 8D method comes from the automotive industry, and today most manufacturing customers require it. The problem is not the method, but how it is used. In our projects, we keep seeing the same pattern: the quality engineer picks up the complaint, fills out the report alone in one afternoon, writes “operator inattention, retrained” in D4, and sends it off. Three months later, the same complaint comes back.

Where did it go wrong? The steps were filled in, but none was verified. Nobody appointed a team. Nobody set deadlines, and the change never made it into the FMEA or the control plan. All that remained of a project management method was paperwork.

A form can be filled out in one afternoon. Only a project eliminates the cause.

Eight steps and three places where they get cut short

The eight steps are a given. Only their content gets cut short, in three weak spots: the never-ending D3, the unproven D4, and the skipped D7.

D1 to D3: team, description, containing the damage

In D1, put together a team of 4–6 people: quality engineer, process engineer, operator, maintenance technician, and logistics if needed. One person is not a team. In D2, describe the problem so everyone understands it without a phone call: which part, which parameter, how many pieces, since when, where the defect was found. For example: “32H7 hole out of tolerance by 0.04 mm, 38 pcs out of 12,400, first occurrence in week 23, found at the customer's assembly”.

D3 means temporary containment within 24 hours: sorting stock and current production, 100% inspection, and, if necessary, replacing parts at the customer's. They protect the customer while you look for the cause. Watch out for the word “temporary”. When the sorting lasts half a year, you never saw D5 and D6 through.

D4: a cause you can prove

Here we build on the second lesson: the 5 Whys and Ishikawa. The yardstick is simple. You must be able to verify the cause — you can reproduce the defect from it on the shop floor, and once the cause is removed, it disappears. If you write “human error” in D4, you have not found the cause. Human error is a symptom asking why the process let one person do harm.

D5 to D8: solution, verification, prevention, closure

In D5, choose actions that attack the cause, not the symptom. In D6, implement them and verify effectiveness with data: 30 days of dimension monitoring, a Cpk calculation, NOK parts per shift. D7 asks where else the same cause lives — the same fixtures on other lines, similar processes, the FMEA, control plan, TPM intervals, work instructions. Companies skip it most often — yet precisely this step keeps the defect from recurring elsewhere. D8 closes the project: evaluation, thanks to the team, lessons learned.

An example from machining: a complaint about 38 pieces

An automotive supplier machining aluminum castings, 180 employees. The customer complained about a 0.04 mm deviation of the 32H7 hole. Team in D1: process engineer, operator, maintenance technician, quality engineer. Within 24 hours in D3, they sorted 12,400 pcs of stock and three days of production, and the customer received replacements from an inspected batch.

In D4, the 5 Whys led from the deviation, through a shifted part, to a worn centering pin in the clamping fixture — and to the finding that the fixture had no defined pin replacement interval. Verification was unambiguous: with a new pin, tolerances returned; with a worn one, the defect was reproduced on test parts. Cause confirmed, not a hypothesis.

D5 to D7: pins with a 20,000-cycle service life, a dimension check on the first piece of every shift, intervals in TPM, an updated FMEA and control plan. After one year: zero repeat complaints for that fixture and the complaint resolution cycle shortened from 24 to 9 days.

How to give 8D the rules of a project

If 8D is to be a project, it needs a minimum: a team leader, a short daily meeting (15 minutes), and milestones. Deadlines that have worked well for us: D3 within 24 hours, D4 within 5 working days, D6 within 30 days, the whole report within 6 weeks. Evaluate effectiveness with data 30 and 90 days after closure — ppm, Cpk, NOK parts per shift.

Second principle: do not close an 8D when the report is sent, but when its effectiveness is verified. A report without verification is a promise, not a result.

Exercise for this week: take the last 8D report you closed and go through it with three questions.

1. Is the cause in D4 proven — can you reproduce the defect from it?
2. Did D7 change any system document: the FMEA, control plan, TPM, a work instruction?
3. How many days did the sorting in D3 last, and when did someone end it?

If at least two answers fail, what you have on the table is a form, not a project. Review the results with the team for 20 minutes and propose one change right away.

A complaint is a lesson the customer paid for. A good 8D extracts the maximum for the entire company. If you want to practice 8D on your own complaints with our consultants, see our e-learning offer. Next time, PDCA — the cycle that turns a one-off solution into lasting improvement.

Quiz — test your knowledge