Cynefin Framework
Core principle: Before applying any solution, classify the problem domain. The right action in one domain is the wrong action in another. Misclassifying the domain is one of the most common causes of failed interventions.
The Five Domains
1. 🟢 Clear (formerly "Simple")
Relationship: Cause and effect are obvious to everyone. Nature: Best practices exist and are well-known. Approach: Sense → Categorize → Respond
- Apply the known best practice
- Automate where possible
- Don't over-complicate it
Signs you're here: The answer is obvious. Anyone competent would do the same thing.
Danger: Complacency. Clear domains can suddenly become chaotic if conditions change and nobody noticed.
Examples: Resetting a password, deploying a known-working config, following a checklist.
2. 🔵 Complicated
Relationship: Cause and effect exist but require analysis or expertise to see. Nature: Multiple right answers exist; good practice (not best practice). Approach: Sense → Analyze → Respond
- Bring in experts or do analysis
- Evaluate trade-offs between valid options
- Allow time for assessment before acting
Signs you're here: You need an expert, but once they analyze it, the answer becomes clear. There's a knowable right answer.
Danger: Expert overconfidence. Experts can mistake complicated for clear and stop questioning assumptions.
Examples: Software architecture decisions, performance tuning, medical diagnosis, legal strategy.
3. 🟡 Complex
Relationship: Cause and effect are only visible in retrospect. You can't predict outcomes. Nature: Emergent practice — no "right answer" exists in advance. Approach: Probe → Sense → Respond
- Run safe-to-fail experiments (small, reversible, parallel)
- Amplify what works, dampen what doesn't
- Expect and embrace emergence
- Don't plan for a single outcome
Signs you're here: Experts disagree. Past data is unreliable. The system involves human behavior, social dynamics, or high interconnectedness.
Danger: Applying complicated or clear approaches (best practices, expert analysis, big-bang solutions) to complex problems. This is the most common mistake.
Examples: Product-market fit discovery, organizational culture change, user behavior patterns, AI system behavior, team dynamics.
4. 🔴 Chaotic
Relationship: No discernible cause and effect. The system is in crisis. Nature: Novel practice — act first to stabilize. Approach: Act → Sense → Respond
- Take immediate action to reduce harm and establish order
- Don't wait for analysis — the situation is changing faster than analysis can help
- Move to complex or complicated once stability is achieved
Signs you're here: Crisis. System is actively failing. No time for deliberation.
Danger: Staying in chaotic mode too long. Once the crisis is contained, move to a more deliberate approach.
Examples: Production outage, security breach, public crisis, team breakdown.
5. ⚫ Disorder (the center)
Relationship: Unknown. You don't know which domain you're in. Approach: Break the situation into parts and classify each one.
- Multiple people will interpret the situation through their own domain lens (causing conflict)
- Get clarity before acting
Output Format
🗺️ Domain Classification
- Assessed domain: Clear / Complicated / Complex / Chaotic / Disorder
- Confidence: High / Medium / Low
- Key signals that led to this classification
- Alternative interpretation: Could this be misclassified? What would that imply?
⚠️ Domain Mismatch Check
Is the user (or current plan) applying an approach from the wrong domain?
| Current approach | Actual domain | Mismatch risk |
|---|---|---|
| Best practice playbook | Complex | High — will fail |
| Big analysis, single solution | Chaotic | High — too slow |
| Experiments / probing | Clear | Waste of time |
🎯 Recommended Approach
Based on the domain:
- Clear: Apply best practice [name it]. Consider automating.
- Complicated: Engage expert analysis. Evaluate trade-offs between [option A] and [option B].
- Complex: Design 2–3 safe-to-fail experiments. Define what "amplify" and "dampen" signals look like.
- Chaotic: Act immediately on [specific stabilizing action]. Reassess domain once stable.
🔁 Domain Transitions to Watch
- Is this domain shifting? (e.g., a complicated situation becoming complex due to new uncertainty)
- What would cause the situation to move into chaos?
- What's the path toward clear once this is resolved?
Practical Heuristics
- If experts disagree → probably Complex, not Complicated
- If it involves people and behavior at scale → probably Complex
- If it's urgent and breaking → probably Chaotic; act first
- If there's a known playbook → probably Clear or Complicated
- If past solutions stopped working → domain may have shifted
The Most Common Mistake
Treating a Complex problem as Complicated: hiring a consultant to define "the right answer" and then executing a big plan — when the system is actually emergent and unpredictable. The result: the plan doesn't work, more consultants are hired, the plan gets bigger, it still doesn't work.
Fix: Run small experiments instead. Let patterns emerge. Amplify what works.