Four ways to structure the same question
Question: Why did a subscription product lose profit?
Equation: Profit = revenue − cost; revenue = customers × price; customers = new customers + retained customers.
Customer journey: Awareness → trial → activation → retention → expansion.
Process: Product development → acquisition → onboarding → service → renewal.
Hypothesis: Profit fell because churn increased after a price change.
How to choose
Use the structure that most directly exposes the decision driver. The best framework isn't the one with the most branches; it is the one that makes the next analysis obvious.
Frameworks are ways to decompose reality
A framework is useful when it makes the next analysis clearer. Memorized lists often fail because they aren't derived from the decision. Build the structure from equations, processes, stakeholders, customer journeys, capabilities, or hypotheses.
Mathematical decomposition
Profit = revenue − cost. Revenue can be price × volume, and volume can be customers × frequency × units. Cost can be fixed plus variable. The equation reveals which data can explain the result.
Process decomposition
A process structure follows stages: product development, procurement, production, distribution, sales, service, and renewal. It is useful for operations, bottlenecks, quality, and customer experience.
Customer-journey decomposition
Awareness, consideration, acquisition, onboarding, usage, retention, and expansion can explain growth and customer problems. The exact stages should match the business.
Stakeholder decomposition
Customers, employees, suppliers, regulators, owners, and partners can reveal different incentives and constraints. This is useful when adoption or policy matters.
Capability decomposition
Strategy, people, process, technology, data, governance, and capital can explain whether an organization can execute a plan.
Hypothesis tree
Begin with a provisional answer and break it into conditions. “Enter the market” may require attractive demand, a credible advantage, acceptable economics, and manageable execution risk. Each condition can be tested.
MECE as a diagnostic
Branches should avoid unnecessary overlap and cover the important possibilities. Perfect MECE isn't always possible or useful. The practical goal is to prevent double counting and missing a decisive issue.
Choosing the structure
Use the structure that maps most directly to the decision and available evidence. A profitability equation is better than a generic “market, company, competition” list when the question is why margin fell. A process map is better when the problem is delay or quality.
Converting analysis to a pyramid
After the investigation, reverse direction. State the answer, group the reasons, and place evidence under each. The final communication shouldn't reproduce the entire issue tree.