MECE Issue Tree
Use $ARGUMENTS as initial context.
When to use this skill
- Diagnosing root causes in performance decline or execution failures.
- Structuring strategic questions into mutually exclusive branches.
- Creating a prioritized analysis plan before data deep-dives.
- Aligning teams on problem scope and ownership.
When not to use this skill
- Do not use it as a final recommendation, market forecast, or execution plan.
- Do not use it when the problem is already structured and only a falsifiable test plan is missing.
Required inputs
- Problem statement, metric, and baseline.
- Scope boundaries (segment, geography, time horizon).
- Available data and decision deadline.
- Metric definition, numerator/denominator, units, and known exclusions.
Workflow
- Convert the request into one decision-oriented problem statement.
- Select tree type: driver, process, option, or hypothesis tree.
- Build 2-3 levels of MECE branches with parallel labels.
- Run formal checks for overlap, gaps, level-mixing, units, and parent-child reconciliation.
- Prioritize branches by impact, controllability, and learning speed.
- Translate top branches into an analysis backlog with owners and timing.
Ask-first questions
Ask up to 3 questions before building the tree:
- Which metric and baseline define the problem severity?
- What scope is explicitly in or out?
- What decision must this tree support?
Assumption policy
- Proceed if data is incomplete, but list assumptions in a dedicated section.
- Tag assumptions with confidence and validation path.
- Do not invent branch evidence; flag unknowns explicitly.
- State whether the tree is a driver, process, option, or hypothesis structure and why.
Output contract
Always produce these sections in order:
- Context
- Decision or Recommendation
- Analysis
- Risks
- Next Actions
- Assumptions
- Every backlog item includes an owner, due date, required dataset, and success signal.
- Label evidence as Fact, Inference, Assumption, or Unknown.
Guardrails
- No branch overlap at the same level.
- No mixing causes and outcomes in one branch layer.
- No "other" bucket unless unavoidable and quantified.
- Keep branch naming at equivalent abstraction depth.
- Do not mix causes, outcomes, actions, and metrics at the same branch level.
Handoffs
- Use
consulting-hypothesis-driven-80-20 when the top branches need falsifiable tests and kill criteria.
- Use
decision-analysis-under-uncertainty when the tree must become an explicit option decision.
- Use
pyramid-principle-structured-communication for the final executive storyline.
Resources
references/issue-tree-patterns.md - Tree patterns and branch design rules.
references/mece-checks.md - Validation gates and failure diagnostics.
templates/issue-tree.md - Decision-ready tree template.
examples/issue-tree-example.md - Golden example with partial information.
Keywords
issue tree, MECE, root cause, problem structuring, analysis backlog, driver tree
1---2name: consulting-issue-tree-mece3description: Build MECE issue trees for complex business problems. Use when you need rigorous problem decomposition, branch prioritization, and a decision-ready analysis backlog.4---56# MECE Issue Tree78Use $ARGUMENTS as initial context.910## When to use this skill11- Diagnosing root causes in performance decline or execution failures.12- Structuring strategic questions into mutually exclusive branches.13- Creating a prioritized analysis plan before data deep-dives.14- Aligning teams on problem scope and ownership.1516## When not to use this skill17- Do not use it as a final recommendation, market forecast, or execution plan.18- Do not use it when the problem is already structured and only a falsifiable test plan is missing.1920## Required inputs21- Problem statement, metric, and baseline.22- Scope boundaries (segment, geography, time horizon).23- Available data and decision deadline.24- Metric definition, numerator/denominator, units, and known exclusions.2526## Workflow271. Convert the request into one decision-oriented problem statement.282. Select tree type: driver, process, option, or hypothesis tree.293. Build 2-3 levels of MECE branches with parallel labels.304. Run formal checks for overlap, gaps, level-mixing, units, and parent-child reconciliation.315. Prioritize branches by impact, controllability, and learning speed.326. Translate top branches into an analysis backlog with owners and timing.3334## Ask-first questions35Ask up to 3 questions before building the tree:361. Which metric and baseline define the problem severity?372. What scope is explicitly in or out?383. What decision must this tree support?3940## Assumption policy41- Proceed if data is incomplete, but list assumptions in a dedicated section.42- Tag assumptions with confidence and validation path.43- Do not invent branch evidence; flag unknowns explicitly.44- State whether the tree is a driver, process, option, or hypothesis structure and why.4546## Output contract47Always produce these sections in order:481. Context492. Decision or Recommendation503. Analysis514. Risks525. Next Actions536. Assumptions54- Every backlog item includes an owner, due date, required dataset, and success signal.55- Label evidence as Fact, Inference, Assumption, or Unknown.5657## Guardrails58- No branch overlap at the same level.59- No mixing causes and outcomes in one branch layer.60- No "other" bucket unless unavoidable and quantified.61- Keep branch naming at equivalent abstraction depth.62- Do not mix causes, outcomes, actions, and metrics at the same branch level.6364## Handoffs65- Use `consulting-hypothesis-driven-80-20` when the top branches need falsifiable tests and kill criteria.66- Use `decision-analysis-under-uncertainty` when the tree must become an explicit option decision.67- Use `pyramid-principle-structured-communication` for the final executive storyline.6869## Resources70- `references/issue-tree-patterns.md` - Tree patterns and branch design rules.71- `references/mece-checks.md` - Validation gates and failure diagnostics.72- `templates/issue-tree.md` - Decision-ready tree template.73- `examples/issue-tree-example.md` - Golden example with partial information.7475## Keywords76issue tree, MECE, root cause, problem structuring, analysis backlog, driver tree