The fishbone diagram and the 6Ms — structured brainstorming for root cause
A fishbone keeps a root-cause session from turning into a list of someone's favorite suspects. The 6M categories force you to look everywhere a cause could hide, not just where you expect it.
A fishbone (Ishikawa) diagram organizes the possible causes of a problem along a spine pointing at the effect. In manufacturing the branches are the 6Ms — Man/People, Machine, Method, Material, Measurement, and Mother Nature/Environment. Each branch collects candidate causes within that category, so a team brainstorms systematically and no class of cause gets overlooked. It's a structuring tool for hypotheses, not proof — causes still have to be verified.
The six categories
- Man / People — training, technique, staffing, handovers, human error.
- Machine — equipment wear, calibration, settings, maintenance state.
- Method — the process, procedure, work instructions, sequence.
- Material — incoming material properties, lot variation, storage, contamination.
- Measurement — gauges, inspection method, sampling, the measurement system itself.
- Mother Nature / Environment — temperature, humidity, vibration, cleanliness.
How to use it well
1 · Write a sharp effect first
The head of the fish is the problem statement. Vague in, vague out — "quality issues" produces a useless diagram; "12% short shots on cavity 3 since the resin lot change" focuses every branch.
2 · Populate every branch before judging
The discipline is breadth. Force at least one candidate under each M before you start eliminating — that's what stops the team from anchoring on the usual suspect and skipping Measurement entirely.
3 · Drill with why on each twig
Each candidate cause can be pushed deeper with a quick 5 Whys. The fishbone gives breadth; the whys give depth.
4 · The diagram lists hypotheses — now verify
A fishbone never proves anything; it organizes what to test. Rank the candidates and verify the top ones with data before committing to a fix.
Common mistakes
- Listing symptoms as causes. "Parts are bad" is the effect restated, not a cause. Each twig should be something you could change.
- Skipping Measurement. Teams jump to Machine and Material and forget the defect might be a gauge problem — verify the measurement system before chasing the process.
- Treating the fullest branch as the answer. The branch with the most sticky notes is the loudest, not necessarily the cause. Verify, don't vote.
- Stopping at the diagram. A fishbone with no follow-up verification is a wall decoration.
Where this gets slow by hand
Drawing the fishbone is fast; verifying its branches is not. Confirming whether Material, Machine, or Measurement is the real driver means correlating the defect against incoming-lot data, equipment history, and gauge studies — pulling three data sources into one timeline for every candidate cause.
From a branch of hypotheses to a ranked, verified cause
Niobia takes the candidate causes a fishbone surfaces and tests them against the data — correlating the defect with incoming-material lots, equipment and maintenance history, process drift from SPC, and the measurement system itself. Instead of a wall of equally-weighted sticky notes, you get the candidates ranked by how strongly the evidence supports them, with the weak ones ruled out. It fits naturally into the 8D root-cause step (D4).
Frequently asked
What are the 6 Ms in a fishbone diagram?
Man (or People), Machine, Method, Material, Measurement, and Mother Nature (or Environment). They are the six standard categories used to organize candidate causes in a manufacturing Ishikawa diagram so that no class of cause is overlooked.
What is a fishbone diagram used for?
It's used to brainstorm and organize the possible causes of a problem in a structured way. The effect goes at the head and candidate causes are grouped along branches — the 6Ms in manufacturing — so a team explores every category systematically rather than fixating on one.
Is a fishbone diagram the same as root cause analysis?
No. A fishbone organizes and displays candidate causes; it does not prove which is the actual root cause. It's one tool within root cause analysis, and the leading candidates still have to be verified with data before a corrective action is chosen.