An AI coach based on MECE and Barbara Minto's Pyramid Principle. Give Claude a vague or overwhelming problem, get a complete Issue Tree back. With hypotheses per branch, MECE check, and 80/20 prioritisation.
Is the drop due to fewer customers?
High impactLower average order value per customer?
Impact mediaHigher churn of existing customers?
High impactEnd of season clearance: what's the timing?
Impact layer90%: A great deal of time is wasted on unstructured brainstorming when tackling a major issue. With MECE, you first break the problem down into sub-solutions that do not overlap but together cover the whole. Only then do you come up with solutions.
Instead of hours of brainstorming, you'll get an Issue Tree in 10 minutes that shows which sub-question needs to be answered first.
The skill ALWAYS checks whether branches are mutually exclusive (no overlap) and collectively exhaustive (together covering 100% of the problem). Standard McKinsey approach.
Each branch holds a falsifiable hypothesis. You don't collect data without a purpose; you test hypotheses. A proven approach from consulting.
Every analysis concludes with a single action, taking a maximum of one working day. No 30-page reports that disappear into a drawer.
Four fixed phases. First, structure; then, solutions can follow. The order is important: without a MECE Issue Tree, you'll start solving the wrong problem.
Your problem will first be rewritten into a Situation, Complication, and Resolution question. The vague becomes sharp.
The root question is divided into 2-5 branches. Algebraically, process-wise, stakeholder or segment-wise depending on the type of issue.
No vague analysis. Each branch will receive a falsifiable statement that you can later test with data.
The skill highlights which 20% branches are expected to yield 80% in the response. That’s where you start.
SCR formula: situation now, what disturbs it, and what question needs to be answered.
The main question is broken down into sub-questions that do not overlap and together cover 100%.
Each branch has a verifiable statement with impact, required evidence, and owner.
The 80/20 branches are marked and the skill provides one concrete action of a maximum of one working day.
This is what complete output looks like. You give the skill a problem in one sentence, this is what you get back.
Resolution question: why has monthly revenue dropped from 80k to 65k over the past three months?
Hypothesis: the number of new customers per month has decreased because we have less advertising budget.
Hypothesis: leads have fallen by 30% since Q1.
Hypothesis: Conversion rate has dropped due to poorer intake.
Hypothesis: Existing customers are buying less per order due to the price increase.
Hypothesis: existing customers are cancelling due to a better competitor.
The difference between an hour of unstructured brainstorming and ten minutes with an Issue Tree.
Do you want to reverse engineer the skill yourself or use it as a prompt? This is the core of what the skill does behind the scenes.
You are a MECE Problem Structuring Coach trained on the McKinsey method and the work of Barbara Minto.
Reframe the problem as a Situation, Complication, and Resolution question. The resolution question will be the root of the Issue Tree.
Break down the root cause into 2 to 5 sub-questions. Choose the appropriate MECE structure: algebraic, process-based, stakeholder, segment, 7S, or internal x external.
Each branch has a falsifiable statement with impact (high/medium/low), required evidence, and owner.
Test MECE explicitly. Mark 80/20 branches. End with 1 concrete action of maximum 1 working day.
MECE stands for Mutually Exclusive, Collectively Exhaustive. It was introduced by Barbara Minto at McKinsey in the 1960s. It works because it prevents two very common mistakes: double work (overlap between branches) and blind spots (missing branches). With MECE, you can be sure that you cover everything without double-analysing.
For any large, vague or overwhelming issue. Strategic decisions, revenue declines, customer retention problems, productivity questions, market research, organisational issues. Not for pure fact-finding questions or for problems that take less than 2 hours to resolve.
The Pyramid Principle (also by Minto) is how you communicate your answer: top-down, governing thought, supported by arguments. SCQA is a narrative structure for the opening: situation, complication, question, answer. MECE is underneath that: how you break down the problem itself before formulating an answer. They reinforce each other.
No. The skill guides you step by step. You enter a problem in plain language, the skill structures it according to the McKinsey method and explains why the branches are MECE. You don't need to have read Minto, Rasiel or Conn.
Yes. The skill is a standard Claude skill (SKILL.md format). You can use it in Claude Desktop, Claude Code, Claude API, and in Claude for Chrome. Download the zip, unpack it into your skills folder.
Three standard works. Barbara Minto, The Pyramid Principle (1973). Ethan Rasiel, The McKinsey Way (1999). Charles Conn and Robert McLean, Bulletproof Problem Solving (2018). All in the bibliography at the bottom of SKILL.md.
Yes, MIT licence. Open SKILL.md, add domain-specific MECE structures (for example, for your own industry), save as a new skill. Please provide feedback on what you have adjusted, so everyone improves.
Free, MIT licence, no registration. Unpack the zip in your Claude skills map and activate per project.
SKILL.md with the complete McKinsey method approach. Ready for Claude Desktop, Claude Code and Claude API.
Download zip (4 KB)Structuring using MECE is stronger when combined with other frameworks. Three suitable skills from TheSEO.
Communicate your MECE findings top-down as Barbara Minto intended: governing thought first, then supporting details.
View skillOnce the correct MECE branch has been chosen, use the 5 Whys to drill down to the root cause within that branch.
View skillBefore you carry out your 80/20 action: a pre-mortem shows how it can go wrong and what you need to prepare.
View skill