strategy-consulting

name: strategy-consulting description: McKinsey-style strategic analysis using multi-agent consulting workflow. Applies SCQA framing, Issue Trees, MECE decomposition, and Pyramid Principle to produce executive-ready recommendations. allowed-tools: Read, Write, Edit, Glob, Grep, WebSearch, WebFetch, Task

Strategy Consulting Skill

McKinsey-style strategic analysis using a multi-agent consulting workflow.

Overview

This skill applies professional consulting methodologies to strategic questions, producing executive-ready analysis and recommendations.

Philosophy

Structure before content - Frame the problem correctly before diving into analysis.

Hypothesis-driven - Start with a point of view, then validate or refute with evidence.

So what? - Every insight must answer "why does this matter?"

Pyramid Principle - Lead with conclusions, support with evidence.

When to Use

Invoke this skill when you need to:

• Analyze a strategic question systematically
• Build a business case for a decision
• Evaluate multiple options with tradeoffs
• Produce executive-ready recommendations
• Structure complex problems into manageable pieces

Quick Start

"Use strategy consulting to analyze [question]"

or

"/strategy-consulting"

The Seven-Step Workflow

Step 1: Problem Definition (Engagement Manager)

Frame the question using SCQA:

• Situation: What's the current state?
• Complication: What changed or is problematic?
• Question: What specifically are we trying to answer?
• Answer: What's our Day 1 hypothesis?

Output:

## Problem Definition

**Situation**: [Current state]

**Complication**: [What changed/problem]

**Question**: [Specific question to answer]

**Day 1 Hypothesis**: [Initial point of view to test]

Step 2: Scope and Boundaries

Define what's in and out of scope:

• What decisions does this analysis inform?
• What decisions are explicitly out of scope?
• What constraints exist?
• Who are the stakeholders?

Step 3: Issue Tree (Framework Strategist)

Build MECE decomposition of the question:

## Issue Tree

**Core Question**: [From Step 1]

### Branch 1: [First MECE branch]
- Sub-question 1.1
- Sub-question 1.2
- Sub-question 1.3

### Branch 2: [Second MECE branch]
- Sub-question 2.1
- Sub-question 2.2

### Branch 3: [Third MECE branch]
- Sub-question 3.1
- Sub-question 3.2

MECE = Mutually Exclusive, Collectively Exhaustive

• No overlap between branches
• All possibilities covered

Step 4: Quality Check (Partner Critic)

Red-team the framing:

• Is the question the right question?
• Is the hypothesis falsifiable?
• Are the branches truly MECE?
• What's missing?

Step 5: Evidence Gathering (Analyst)

For each branch of the issue tree:

• Gather relevant data
• Apply Fermi estimation where data is missing
• Assess source credibility
• Identify the vital few factors (80/20)

Evidence standards:

• Explicit confidence levels
• Source citations
• Assumptions stated
• Ranges, not false precision

Step 6: Quality Check 2 (Partner Critic)

Validate the analysis:

• Is the evidence sufficient?
• Are conclusions supported?
• What are the key risks?
• What would change our mind?

Step 7: Output Generation (Executive Storyteller)

Produce final deliverables using Pyramid Principle:

Structure:

1. Lead with the answer
2. Support with key arguments
3. Back up with evidence
4. Address alternatives considered

Output Formats

P2 Memo (Default)

# [Title]: [Action-Oriented Headline]

## The Bottom Line

[2-3 sentences: Answer + So What + Recommended Action]

## Key Arguments

### 1. [Argument with action title]

[Evidence and reasoning]

### 2. [Argument with action title]

[Evidence and reasoning]

### 3. [Argument with action title]

[Evidence and reasoning]

## Alternatives Considered

| Option | Pros | Cons | Verdict |
|--------|------|------|---------|
| [Option A] | | | Rejected: [reason] |
| [Option B] | | | Selected |

## Risks and Mitigations

| Risk | Likelihood | Impact | Mitigation |
|------|------------|--------|------------|
| | | | |

## Next Steps

1. [Action] - Owner: [Name] - By: [Date]
2. [Action] - Owner: [Name] - By: [Date]

Deck-Ready Content

For presentation slides:

# Slide 1: Executive Summary
**Headline**: [Action-oriented conclusion]
**Bullets**:
- Key point 1
- Key point 2
- Key point 3

# Slide 2: [Topic]
**Headline**: [Conclusion for this topic]
**Visual**: [Chart/diagram description]
**Talking points**:
- Point 1
- Point 2

Slack Summary

For quick communication:

[One-line answer]

Key points:
• [Point 1]
• [Point 2]
• [Point 3]

Recommendation: [Action]
Link to full analysis: [link]

Agents Used

Mental Models Applied

WWHTBT (What Would Have to Be True)

For each option, identify assumptions that must hold:

For [Option X] to be the right choice:
1. [Assumption 1] would have to be true
2. [Assumption 2] would have to be true
3. [Assumption 3] would have to be true

Then prioritize testing the riskiest assumptions.

Fermi Estimation

When precise data isn't available:

Question: How many X?
= [Factor 1] × [Factor 2] × [Factor 3]
= [Estimate 1] × [Estimate 2] × [Estimate 3]
= [Range: Low - Mid - High]

Confidence: [Level] because [reasoning]

80/20 Analysis

Identify the vital few factors:

Of [total factors], [top 20%] drive [80%] of [outcome].
Focus on: [Factor 1], [Factor 2]
Deprioritize: [Other factors]

Example Workflow

Question: "Should we expand into the healthcare vertical?"

Step 1 - SCQA:

• S: We serve enterprise clients in media and retail
• C: Healthcare vertical showing interest; competitors entering
• Q: Should we prioritize healthcare expansion in 2026?
• A: Yes, but with focused entry through existing client referrals

Step 3 - Issue Tree:

1. Is the market attractive?
   • Market size
   • Growth rate
   • Competitive intensity
2. Can we win?
   • Capability fit
   • Reference customers
   • Competitive differentiation
3. Is it worth it?
   • Investment required
   • Timeline to profitability
   • Opportunity cost

Step 7 - Output: P2 memo with recommendation, supporting evidence, and implementation plan.

Best Practices

1. Start with the answer - Have a hypothesis before diving into analysis
2. Be MECE - Structure thinking to be complete and non-overlapping
3. Show your work - Make reasoning transparent
4. Quantify when possible - Numbers beat adjectives
5. Address alternatives - Show you considered other options
6. Name the risks - Credibility comes from acknowledging uncertainty

Troubleshooting

Analysis going in circles:

• Revisit the problem definition
• Check if the question is answerable
• Narrow the scope

Not enough evidence:

• Use Fermi estimation
• State assumptions explicitly
• Acknowledge uncertainty

Stakeholders disagree:

• Return to SCQA
• Find the root disagreement
• Build shared understanding of the question first