2.3The Ishikawa diagram (cause-effect)
The essentials in 30 seconds
- Ishikawa = breadth: the effect as head on the right, branches per cause families — no path forgotten.
- The 5Ms: Manpower, Machine, Method, Material, Milieu (adaptable, + Management/Measurement).
- Reflex: as a team for exhaustiveness, then dig likely causes with the 5 whys.
Running mini-case — Sara sweeps the 5Ms on her errors
Sara's rework piles up entry and measurement errors. Instead of blaming only training, she draws her fish: Manpower (training?), Method (fuzzy procedure?), Machine (clunky software?), Material (wrong input data?), Milieu (noise, interruptions?). She finds ignored causes in several families — then digs the most likely.
Sequel to cases 2.1-2.2 — Sara moves from the single chain (5 whys) to the full map.
1. An overview of causes
When a problem has multiple causes, the Ishikawa diagram (cause-effect, « fishbone ») offers a structured overview. You place the effect (the problem) at the fish's head, on the right, and a big horizontal spine runs from it.
On this spine, branches graft, each representing a cause family. You detail all possible causes by category, avoiding omissions and focusing on a single path.
2. The 5M method
Families are often structured by the 5M: Manpower (people), Machine (equipment), Method (procedures), Material (inputs, data), Milieu (environment).
You can add a 6th M (Management/Measurement) or adapt the categories to the field. For each family: what causes here could produce the problem? You can dig each with the 5 whys.
3. Exhaustive and collective
Ishikawa's value: it forces an exhaustive and organized analysis (you sweep all categories), reveals causes you hadn't considered, and is built as a team.
It is the reference tool for multi-cause problems, complementary to the 5 whys (which dig one chain). Once causes are mapped, you treat the most likely and impactful first.
🔎 Concrete example — Ishikawa on a high error rate
Too many entry errors: sweep the 5M.
Manpower: insufficient training?
Method: unclear procedure?
Machine: unergonomic software?
Material: wrong input data?
Milieu: noise, interruptions?Result: you discover overlooked causes in several families.
Reflex: then dig the most likely ones with the 5 whys.
Section vocabulary
- Ishikawa diagram
- A « fishbone » cause-effect diagram mapping all cause families of a problem.
Simple example: Sara's fish on her errors, effect on the right, 5Ms as branches.
What it is not: a vote to pick an idea or a schedule.
Reflex: place the effect on the right before any branch.
- Effect (fish head)
- The problem to analyze, placed on the diagram's right.
Simple example: « too many entry errors » as the fish head.
What it is not: an already-decided solution, placed as head.
Reflex: phrase the effect as a precise problem, not a solution.
- Cause family
- A category grouping same-nature causes (one diagram branch).
Simple example: Method: fuzzy procedure, missing checklist.
What it is not: an isolated cause with no category or branch.
Reflex: for each branch: « what, here, could produce the effect? ».
- 5M method
- Classic categories: Manpower, Machine, Method, Material, Milieu (sometimes +Management/Measurement).
Simple example: Sara sweeps the 5Ms and adds Measurement for her checks.
What it is not: five minutes or five mandatory measurements.
Reflex: adapt categories to the field if needed.
- Exhaustive analysis
- Organized sweep of all cause categories, to forget none.
Simple example: all branches filled before prioritizing.
What it is not: focusing on a single family from the start.
Reflex: fill each family before judging likelihoods.
In practice — Mapping causes with Ishikawa
- Place the effect (the problem) on the right (fish head) and draw the main spine.
- Create a branch per cause family (5M: Manpower, Machine, Method, Material, Milieu; adapt).
- For each family, list all possible causes (as a team); dig with the 5 whys.
- Identify the most likely/impactful causes to treat first.
Ready-to-say script — launch a team Ishikawa (1 min)
« The effect is on the right: too many entry errors. We fill each branch — people, method, machine, data, environment — no judging. Then we circle the two most likely causes and dig them. »
Say at the whiteboard: breadth first (all paths), depth after.
Copy-paste template — Ishikawa 5M grid
EFFECT (head, on the right): _______________
MANPOWER: _______________
MACHINE: _______________
METHOD: _______________
MATERIAL: _______________
MILIEU: _______________
2 CAUSES TO DIG (5 whys): _______________Small business / tight budget
On an A3 sheet: effect on the right, five hand-drawn branches — fifteen minutes suffice.
Algerian SME
Noisy workshop, outages, variable suppliers: fill Milieu and Material first, often decisive here.
Service / retail
Adapt branches: People, Procedures, Customers, Technology, Environment instead of industrial 5Ms.
Project team
First brainstorm all causes loose, then group them into families afterward.
✓ Best practices
- Build it as a team: each one lights up a cause family.
- Adapt the categories to your field (the 5M aren't sacred).
- Then dig the likely causes with the 5 whys (breadth + depth).
✗ Common pitfalls
- Focusing on a single family and forgetting the others.
- Staying on the surface without digging the retained causes.
- Confusing likely causes with proven ones (verify afterward).
Key takeaways
- The Ishikawa diagram maps all cause families of a problem.
- The effect (the problem) is on the right; branches represent cause families.
- Classic families (5M): Manpower, Machine, Method, Material, Milieu (+Management/Measurement).
- It forces an exhaustive analysis, is built as a team, and complements the 5 whys.
48h Challenge — apply tomorrow
Draw the fish of 1 real problem (template above): effect on the right, 5 filled branches as a team or solo. Circle 2 likely causes to dig.
- I place the effect and fill the 5 branches.
- I tell likely and proven causes apart.
- I did the 48h challenge or planned its date.
Frequently asked questions
Ishikawa or 5 whys?
The 5 whys follow a single linear chain: ideal to dig deep into a main cause, fast and tool-free. Ishikawa explores all causes in parallel, by families: ideal for a complex multi-cause problem, to sweep all paths, or as a team. They are mostly complementary: build an Ishikawa first (broad view of all families), then apply the 5 whys on the likely causes (depth). Ishikawa for breadth, 5 whys for depth.
Must you use the 5M categories?
No: the 5M are a classic starting point (from industry), to adapt to your problem. There's the 6M (with Management or Measurement), or field-specific categories (for a service: People, Procedures, Customers, Technology…). The key is to have relevant categories that help explore all dimensions without omission. You can also brainstorm all causes then group them into families afterward. The tool is flexible: what counts is the broad, organized sweep.
What does the Ishikawa diagram enable?