The situation

The dashboard bezels are warping. In the meeting, tooling blames the material, the material supplier blames the process, and the process team blames the night shift. Each person has a theory. Nobody has a list.

A fishbone diagram puts all theories on one page, sorted, so that you can check them one by one instead of arguing about them.

What a fishbone is for

A fishbone diagram is a picture of possible causes of one problem. The problem sits at the “head” on the right. A long line, the spine, runs to it. Branches off the spine are categories. Causes hang from the categories.

It is a thinking aid for a team. It does not prove anything. It makes sure the team looks in every direction before choosing one.

Terms used here:

  • 6M: six common categories for manufacturing: Machine, Method, Material, Manpower, Measurement, and Milieu (environment).
  • Hypothesis: a possible cause that nobody has checked yet.
  • Gemba: the real place where the work happens.

A short history

The diagram is named after Kaoru Ishikawa. Sources say he first used it in 1943, at Kawasaki Steel Works, and popularised it in Japanese quality circles from the 1960s. It is counted among the seven basic tools of quality. We could not check the 1943 detail against a primary source, so treat the exact year as reported, not proven.

Where it applies, and where it does not

Use a fishbone when:

  • the cause is unknown and several things could be involved;
  • a team must look at machine, method, material, and people together;
  • you need the D4 step of an 8D or the cause box of an A3;
  • departments are pointing at each other.

Do not use it when:

  • the cause is already known. A diagram for an empty oil tank is theatre;
  • you need numbers. The diagram sorts ideas. It cannot replace measurement or designed experiments;
  • you need to track actions. A fishbone shows causes, not who does what.

How to do it

The example is fictional. Mazovia Precision Plastics moulds dashboard bezel BZL-092. Grzegorz is a quality technician. Marta is the tooling engineer. Piotr leads the shift.

  1. State the effect. Write one measurable sentence at the head. Warpage and sink marks above 0.8 mm on the mounting tabs.
  2. Draw the spine and the main bones. Use the 6M as a start. Rename them if your process needs it.
  3. Brainstorm causes on each bone. Everyone adds ideas. No arguing yet. The team lists these:
    • Machine: cooling channel 3 is restricted by scale.
    • Method: mould temperature setpoint was lowered from 85 °C to 70 °C to speed up the cycle.
    • Material: regrind ratio is 35%, above the 15% limit.
    • Measurement: a caliper may flex on soft parts.
    • Milieu: chiller water warms on hot afternoons.
    • Manpower: the night setup technician may not know the new recipe lock.
  4. Go deeper on promising causes. Ask “why does this happen?” and add a sub-branch. Why is regrind at 35%? The feeder calibration was wrong.
  5. Go and see. Check each cause on the floor. Measure the flow, read the setpoint, weigh the blend. Mark each one as confirmed or ruled out. A coordinate measuring machine shows the real warpage is 1.2 mm, so the caliper was reading low. That is a finding too.
  6. Pick the verified few and dig further. Take the two or three confirmed causes into a 5-Why. Here the cooling restriction, the unauthorised setpoint change, and the regrind feeder all go forward.

Three ways it goes wrong

1. The “Manpower” dumping ground. The people bone fills up with “did not pay attention”, while Machine and Method stay empty. That is blame wearing a diagram. “Operator error” is not a root cause. If you write a human cause, ask what in the setup, training, or workload let it happen.

2. Brainstorm treated as proof. Sixty sticky notes feel like work. They are only guesses. Without checks at the gemba, opinions walk straight into the action list. Evidence over opinion.

3. A diagram in a drawer. Someone draws it to satisfy an audit and files it. No bone is tied to a fix or a later check. If a cause is confirmed, it needs an action, an owner, and a date.

Our recommendations

  • Draw it with the people who run the process. They know where the oddities are.
  • Write causes as things, not people. “Cooling channel 3 restricted” is checkable. “Careless setup” is not.
  • Tag every cause. Open, confirmed, or ruled out. Keep the ruled-out ones visible. They stop the same debate next month.
  • Add a sub-branch only when a cause has a reason behind it. Do not draw for its own sake.
  • Do not stop at the diagram. A fishbone starts the investigation. It does not end it.
  • Use a plain paper first. A whiteboard works well. Draw it digitally when you need to share or print.

Try it

Our fishbone tool is coming soon. It will run in your browser, with no account, and your data will stay on your device. Meanwhile, see the tool page: /tools/fishbone.

From the book

This guide follows Chapter 8, the fishbone chapter, of the field manual. See /book for the book.