Lean teams use it to brainstorm causes widely before narrowing down. It keeps a group from fixating on the first idea and shows where to collect data.
Where it comes from
The diagram was developed by Japanese quality expert Kaoru Ishikawa, which is why it is also called an Ishikawa diagram. It is often listed among the basic tools of quality control.
At a glance
A fishbone for uneven paint thickness
Swipe sideways to see the whole diagram.
How to apply Fishbone Diagram
- 1
Write the problem at the head
State the effect clearly and specifically. Draw a horizontal spine leading to it.
- 2
Choose cause categories
Common sets are the 6Ms: people, machine, method, material, measurement and environment. Office teams often use process, policy, people and systems.
- 3
Brainstorm causes
Have the team add possible causes under each category. Ask why to add sub-causes as smaller bones.
- 4
Select likely causes
Mark the causes the team thinks matter most. Base the choice on evidence where you have it.
- 5
Verify with data
Test the selected causes at the gemba or with data. Drill into confirmed ones with 5 Whys.
Worked example (illustrative)
A coating line sees paint thickness vary by up to 18 microns across a panel, against a limit of 10. The team lists causes under method, machine, material and environment, including spray nozzle wear and humidity swings. Data shows the spread is about 6 microns on nozzles under 200 hours old and about 16 on nozzles over 400, while humidity shows no clear link. Nozzle replacement at 250 hours moves into the PM plan, and the spread settles near 7 microns.
Common mistakes
- Treating every cause on the diagram as proven instead of verifying the likely ones.
- Building the diagram without the people who do the work.
- Writing vague causes such as "training" that cannot be tested or acted on.
When Fishbone Diagram is not the right tool
A fishbone diagram lists possible causes. It does not prove any of them. Use data or tests to confirm the likely causes before acting.
For a simple problem with one clear chain of cause, five whys is quicker.
Choosing the categories
Manufacturing teams often use six categories: people, machine, method, material, measurement and environment. Older versions call people "man" and environment "mother nature".
Service and office teams often use process, policy, people and systems instead. Pick categories that help the team think, then add or drop as needed.
Narrowing the causes
Once the diagram is full, the team votes on the most likely causes. Data, such as a Pareto chart of defect types, then checks the vote.
In the illustrative coating example, the team votes for three likely causes: nozzle wear, humidity and paint viscosity. A month of readings shows nozzle age explains about 70% of the out-of-limit panels, viscosity about 15% and humidity almost none. The team acts on the nozzles first and keeps watching viscosity.
Guides on this topic
- Root cause analysis example: 6 worked samples and a templateA root cause analysis example for six settings, from a machine stoppage to warehouse picking errors, plus root cause statement wording and common mistakes.Read the guide
- Lean vs Six Sigma: differences, history and when to use eachLean vs Six Sigma compared: what each method targets, where they came from, how they differ in manufacturing and how Lean Six Sigma combines them.Read the guide
- How to write an A3 report: step by step, with an exampleHow to write an A3 report section by section, with a worked A3 example and the coaching questions that turn A3 problem solving into learning.Read the guide
- Process improvement methodologies: 9 methods comparedProcess improvement methodologies compared: PDCA, lean, kaizen, Six Sigma, TQM, theory of constraints, BPR and retrospectives, with how to choose and combine.Read the guide
Sources
Books and papers show their publication year. Web pages show the month we last checked them.
- 1.Kaoru Ishikawa, Introduction to Quality Control (3A Corporation, 1990)
- 2.Thomas Pyzdek and Paul A. Keller, The Six Sigma Handbook, 4th edition (McGraw-Hill, 2014)