Problem Solving
Break through complex problems with structured decomposition, root cause analysis, and systematic solution generation.
What This Skill Does
The Challenge: Complex problems resist simple solutions. Teams get stuck in surface-level fixes, miss root causes, or debate symptoms instead of diagnosing systems. Unstructured problem-solving wastes time and produces low-quality decisions.
The Solution: Problem Solving skill applies structured frameworks (5 Whys, MECE decomposition, fishbone analysis, decision matrices) to diagnose root causes, generate solution spaces, and evaluate options systematically.
Activation
Implicit: Activates when user signals being blocked, stuck, or needing a systematic approach to a complex challenge.
Explicit: Activate via prompt:
Activate problem-solving skill to diagnose why our user retention dropped 15% last month
Capabilities
1. Root Cause Analysis
Systematically trace symptoms to underlying causes.
5 Whys template:
Problem: User retention dropped 15%
Why 1: Users aren't returning after onboarding
Why 2: They didn't reach the activation milestone
Why 3: The activation step has a UX blocker
Why 4: We changed the UI without user testing
Why 5: No process for UI change validation
Root cause: Missing change management process
2. MECE Decomposition
Break complex problems into mutually exclusive, collectively exhaustive parts.
Decomposition structure:
Retention problem
├── Acquisition quality (wrong users?)
├── Onboarding experience (too much friction?)
├── Product value (promise not delivered?)
└── External factors (competitor launched feature?)
3. Solution Matrix Evaluation
Score solutions against impact and feasibility criteria.
Scoring matrix:
| Solution | Impact (1-5) | Feasibility (1-5) | Score |
|---|---|---|---|
| Fix UX blocker | 5 | 4 | 20 |
| Add tooltips | 3 | 5 | 15 |
| Redesign onboarding | 5 | 2 | 10 |
4. Decision Documentation
Document problem diagnosis, solution rationale, and next steps.
Decision doc format: Problem → Root cause → Options considered → Decision → Success metrics
Prerequisites
- Clear problem statement with measurable symptoms
- Relevant data or context about the situation
- Stakeholders identified who own the problem
Configuration
No tools required. Works through structured conversational analysis.
Best Practices
1. Define the problem before solving it Resist jumping to solutions. Spend 30% of time on diagnosis, 70% on solution design.
2. Separate symptoms from causes Retention drop is a symptom. Use 5 Whys to find the actual cause before picking a solution.
3. Generate 3+ solutions before evaluating Premature solution narrowing leads to suboptimal choices. Always compare alternatives.
Common Use Cases
Use Case 1: Marketing Metric Decline
Scenario: Conversion rate dropped unexpectedly, unclear why.
Workflow:
- Define symptom with data (CVR was 3.2%, now 2.1%)
- MECE decomposition: traffic quality, landing page, offer, pricing
- Hypothesis generation for each branch
- Data check to validate/eliminate hypotheses
- Root cause identified, solution matrix built
Output: Root cause report with prioritized fix list.
Use Case 2: Team Productivity Block
Scenario: Engineering and marketing misaligned on feature priorities.
Workflow:
- Define the conflict precisely (feature requests vs roadmap capacity)
- Fishbone: Identify process, communication, and incentive misalignments
- Generate process solutions (shared OKRs, weekly sync, prioritization framework)
- Score by ease of adoption and impact
- Recommend minimum viable process change
Output: Alignment process proposal with quick win and long-term fix.
Troubleshooting
Issue: Root cause analysis goes circular Solution: Each “why” must produce a different answer. If repeating, you’ve found the root cause or hit a constraint boundary.
Issue: Team disagrees on root cause Solution: Data-validate each hypothesis. Disagreements on root cause are often disagreements on data interpretation.
Related Skills
- Sequential Thinking - Structured step-by-step analysis
- Brainstorming - Solution ideation and exploration
- Analytics - Data-driven diagnosis
Related Commands
/ckm:problem-solving- Systematic problem analysis/ckm:sequential-thinking- Step-by-step reasoning/ckm:brainstorm- Explore solution space