Engineering, production, governance, cost, communication · insight-0293
Open an executive memo with the decision, the evidence range and the main limit
- Evidence type
- opinion
- Difficulty
- easy
- Roles
- Data analyst, Data scientist
- Industries
- aviation, cross-industry
Summary: situation, action, reason, boundary
Situation
A four-page analysis of an overbooking model starts with data sources and methods, and the recommendation is on page four.
Action
Start with the recommendation and the decision requested, then the expected impact as a range with its basis, the main risk or limit, and what would change the recommendation; move methods to an appendix.
Reason
Busy readers act on the top of the document, so it must contain the decision and its conditions.
Boundary
Model launch decisions, investment requests, analysis summaries for leadership.
Symptom or risk
Executives read the first paragraph, miss the recommendation or its conditions, and decide without the key caveat.
Exceptions
Technical review documents for peers can follow a methods-first structure.
Trade-offs
Leading with a recommendation exposes you to challenge early; that is intended.
Pitfalls
Presenting a point estimate of impact without a range or its assumptions.
Example
Hypothetical: 'Raise overbooking on six routes; expected annual gain 0.4 to 0.9 million on simulated data; main risk is denied boardings in peak weeks.'