You get:
- recency bias (the last option heard)
- affinity bias (options from people you like)
- no weight on what actually matters
- false precision without sensitivity analysis
- analysis paralysis with no clear output
But good decisions are not feelings.
They are structured comparisons of trade-offs.
- Criteria must be explicit and weighted
- Scores reveal hidden assumptions
- Sensitivity analysis shows which criteria really matter
- A matrix forces honesty about trade-offs
Without decision discipline, you confuse motion with progress.
This framework forces AI to think like a strategic analyst, not a sounding board.
Assume the role of a strategic decision analyst, multi-criteria decision-making (MCDM) specialist, and trade-off architect. Your task is to help the user make a clear, defensible decision between multiple options using a weighted decision matrix. Before generating, analyze: - the decision's stakes and reversibility - which criteria are truly independent - where weights might hide emotional preferences - the most likely source of bias in scoring Then generate: 1. A weighted decision matrix with: - All options as columns - All criteria as rows - User-assigned weights (1-10) for each criterion - Scores (1-10) for each option on each criterion - Weighted totals per option 2. Top two contenders highlighted 3. Sensitivity analysis: "If your top criterion were 20% less important, would the answer change?" 4. 2-3 sentences of plain-English interpretation INPUTS: Decision Description: [WHAT ARE YOU CHOOSING BETWEEN?] Options (N options): [LIST OPTIONS] Criteria: [LIST WHAT MATTERS IN THE DECISION] Initial Gut Preference (optional, for bias check): [YOUR CURRENT LEADER] Stakes Level: [LOW / MEDIUM / HIGH / CAREER-DEFINING] RULES: - Weights must sum to context (no single 10 dominates unless justified) - Sensitivity analysis is not optional - Flag any criterion that is actually two criteria combined - Output must include a table and plain English - If two options tie, ask one clarifying question about risk tolerance
- Assign weights BEFORE scoring options to avoid post-hoc rationalization.
- If the matrix says Option A but you feel Option B, re-examine your weights — they may be wrong.
- Use sensitivity analysis to identify which criterion is actually doing the work.
- For high-stakes decisions, run the matrix twice with different weight sets.
- Share the matrix with stakeholders — it depersonalizes disagreement.
Decision Description: Which project management tool should our team of 12 adopt?
Options: Asana, ClickUp, Monday.com, Trello
Criteria: Ease of use, Reporting features, Integration options, Price per user, Customer support
Initial Gut Preference: Asana (familiar)
Stakes Level: Medium (team will use for 2+ years)
This framework improves outcomes by forcing:
- explicit criteria weighting before scoring
- sensitivity analysis to test robustness
- plain-English interpretation of math
- trade-off visibility (no hidden assumptions)
- bias flags for gut feelings
Great decisions don’t eliminate intuition — they test it against structure.
Build Better AI Systems
Subscribe for advanced prompt engineering, AI productivity systems, decision frameworks, and practical strategies for builders and leaders.
Leave a Reply