Fishbone Diagrams
A fishbone diagram (also called Ishikawa or cause-and-effect diagram) visually maps the potential causes of a problem. It's particularly useful when multiple factors might contribute and you need to systematically explore possibilities.
The Structure
┌─────────────────────────────────────────────────┐
│ │
Category 1 ───────────┬──────────────────────────────────┤
│ │ │
│ Cause 1.1 ─────┤ │
│ │ │ │
│ Cause 1.2 ─────┤ PROBLEM
│ │ │
Category 2 ───────────┬──────┤ │
│ │ │ │
│ Cause 2.1 ─────┤ │
│ │ │
│ Cause 2.2 ─────┤ │
│ │
└─────────────────────────────────────────────────┘
The "head" of the fish is the problem. The "bones" are categories of potential causes. Smaller bones are specific causes within each category.
Building the Diagram
Step 1: Define the Problem
Write the problem clearly at the head of the fish. Be specific. "Customer complaints" is too vague. "Customers complaining about delivery time" is better.
Step 2: Choose Categories
Draw major bones for each category of causes. Common category sets:
Manufacturing (6 Ms):
- Man (People)
- Machine (Equipment)
- Method (Process)
- Material (Inputs)
- Measurement
- Mother Nature (Environment)
Services (8 Ps):
- Price
- Promotion
- People
- Processes
- Place/Plant
- Policies
- Procedures
- Product
Or create your own:
- Systems
- People
- Process
- External
- (Whatever fits your context)
Step 3: Brainstorm Causes
For each category, brainstorm potential causes. Add them as smaller bones. Don't evaluate yet, just collect.
Step 4: Dig Deeper
For each cause, ask "why?" Add sub-causes as even smaller bones. This is where the Five Whys technique combines with fishbone diagrams.
Step 5: Analyze
Look for:
- Causes that appear in multiple categories
- Clusters of related causes
- Causes with strong evidence
- Causes that are actionable
Step 6: Prioritize
Not all causes are equal. Use evidence and judgment to identify the most likely root causes. Investigate those first.
Example
Problem: High employee turnover
| Category | Causes |
|---|---|
| People | Poor hiring fit, Inadequate onboarding, Manager quality, Team dynamics |
| Process | Unclear expectations, No career path, Slow decision-making, Too much bureaucracy |
| Compensation | Below market salary, Unclear bonus structure, Limited benefits, No equity |
| Culture | Toxic behaviors tolerated, No recognition, Work-life imbalance, Lack of purpose |
| Environment | Poor tools, Office location, Remote policy, Physical workspace |
| External | Competitor recruiting, Industry trends, Economy, Geographic factors |
After investigation, you might find the primary drivers are "Manager quality" and "No career path," with "Below market salary" as a contributing factor.
When to Use It
- Complex problems with multiple potential causes
- Group brainstorming sessions
- When you need to ensure comprehensive analysis
- When different stakeholders have different theories
- Teaching teams to think systematically about causation
When It Doesn't Work
Too Many Causes
If you generate 50 causes, the diagram becomes unwieldy. Prioritize or group similar causes.
No Investigation
Fishbone diagrams generate hypotheses. They don't prove anything. Each suspected cause needs verification.
Consensus Theater
Sometimes teams use fishbone diagrams to create false agreement. Everyone's theory goes on the diagram, but no one's theory gets tested.
Linear Problems
If the cause is straightforward, a fishbone diagram is overkill. Use Five Whys instead.
Best Practices
Use in Groups
Fishbone diagrams work best as group exercises. Different perspectives surface different causes.
Be Specific
"Communication problems" is too vague. "Status updates not reaching stakeholders" is better.
Include Data
Where possible, attach evidence to causes. "Manager quality" backed by "Exit interviews cite manager in 60% of departures" is more persuasive.
Iterate
The first pass is rarely complete. Revisit after investigation. Add what you've learned. Remove what's been ruled out.
Avoid Blame
Keep the focus on systems and processes, not individuals. "John's errors" isn't useful. "Inadequate training for complex tasks" is useful.
Combining with Other Tools
- Five Whys: For each bone, ask why to find deeper causes
- Pareto Analysis: Prioritize which causes to address first
- TRACE Protocol: Use the fishbone in the "Analyze" phase
A fishbone diagram doesn't solve the problem. It maps the territory where the solution lives.