What to remember
- Scenarios that can be understood together
- The same criteria for every option
- Explicit weightings
- Sensitivity analysis before the choice
Build genuinely comparable options
A scenario describes a coherent logic: target market, value proposition, committed resources, timetable, constraints and expected results.
Start with a credible continuity scenario. Then add two or three distinct strategic choices. Do not favor an option by giving it more detail or more favorable assumptions.
- Same time horizon
- Same indicator definitions
- Same depth of analysis
- Attributed assumptions
- Identical non-negotiable constraints
Choose criteria and make weightings visible
Criteria translate the organization’s strategy. A score is useful only when each criterion has a definition, scale and evidence. A weighting should be presented as a governance choice rather than a mathematical truth.
| Family | Possible questions | Caution |
|---|---|---|
| Value | Revenue, margin, strategic advantage | Do not count the same benefit twice |
| Feasibility | Skills, capacity, timing | Separate difficulty from impossibility |
| Risk | Dependencies, compliance, adoption | Assess impact and exposure |
| Reversibility | Exit cost, commitments | Make sunk costs visible |
| Learning | Information produced by the option | Value reversible stages |
Test sensitivity instead of seeking the perfect figure
Identify the two or three assumptions explaining most of the gap between scenarios. Vary each within a justified range. If a small change reverses the choice, the decision is fragile and needs a learning stage.
Simulations can enrich the comparison, but they cannot compensate for poorly defined distributions or correlations. The method should remain explainable to decision-makers.
- Central assumption
- Justified low and high ranges
- Effect on the ranking
- Switching threshold
- Information to collect before committing
Move from the matrix to the recommendation
Grid helps structure matrices and portfolio choices. Plot puts options into perspective, tests scenarios and helps prepare the synthesis. The final recommendation should still explain the reasoning in accessible language.
A useful conclusion states which option is recommended, why, under which conditions, which risks to accept and which signal would trigger a review.
Frequently asked questions
When is a compromise table preferable?
A compromise table is often more faithful when criteria are difficult to combine or when one criterion is disqualifying.
How should a qualitative criterion be handled?
Define an observable scale, give examples for each level and retain the source or judgment supporting the score.
What is a switching threshold?
It is the value of an assumption at which the preferred scenario changes. It shows where research or vigilance should focus.