Skip to main content
Reasoning Prompts

Weighted Decision Matrix

Score multiple options against weighted criteria to make complex decisions more objective and transparent.

intermediateWorks with any modelReasoning
Prompt
Help me choose between these options using a weighted decision matrix.

**Options**: [OPTION_1], [OPTION_2], [OPTION_3]

**Criteria and weights** (must total 100):
- [CRITERION_1]: [WEIGHT_1]%
- [CRITERION_2]: [WEIGHT_2]%
- [CRITERION_3]: [WEIGHT_3]%
- [CRITERION_4]: [WEIGHT_4]%

**Scoring scale**: 1 (very poor) to 5 (excellent)

For each option, score it on each criterion and explain the score in one sentence. Then calculate the weighted score: (score × weight%) summed across all criteria.

Present the results as a table, then identify the winner and explain whether the scores reveal any non-obvious trade-offs I should consider before deciding.

How to Use

Fill in up to 4–5 options and define the criteria that matter most to your decision. The key step is setting the weights before scoring — if you set weights after seeing scores, you'll unconsciously adjust them to justify the option you already prefer. The matrix is most useful when the winner is close (within 5–10 points), because that signals a genuinely hard trade-off worth thinking through carefully.

Variables

VariableDescription
[OPTION_1/2/3]The alternatives you're choosing between (add more as needed)
[CRITERION_1–4]The factors that matter for this decision (e.g., cost, implementation speed, scalability, team expertise)
[WEIGHT_1–4]Percentage importance for each criterion — must sum to 100

Tips

  • Start by brainstorming criteria with: "What are the 5–7 most important factors for deciding [DECISION]?" before building the matrix.
  • If one option wins on cost but loses on everything else, the matrix is surfacing a real tension — don't dismiss the runner-up without explicitly deciding that cost outweighs the other factors.
  • For major decisions, run the matrix twice: once with your initial weights and once with your weights shifted by ±10% to test how sensitive the outcome is to your assumptions.