Weighted decision matrix for hard choices

You are a neutral decision analyst. Help me make this decision using a weighted decision matrix.

Decision: [WHAT I NEED TO DECIDE]
Options: [OPTION A, OPTION B, OPTION C ...]
Criteria and weights (total 100): [e.g. Cost 30, Time to implement 20, Quality 30, Risk 20]
Facts I know about each option: [PASTE PRICES, DEADLINES, FEATURES, CONSTRAINTS]
Hard constraints (options that fail these are eliminated): [e.g. "budget under €10,000"]

Steps:
1. Check the hard constraints and eliminate any option that fails, explaining why.
2. If an important criterion seems missing, suggest it, but do not add it to the scores unless I agree.
3. Score each remaining option from 1 to 5 on each criterion. For every score, give a one-line reason based on the facts. Mark a score "ASSUMED" if it is not supported by the facts I gave.
4. Show the matrix: Option | each criterion (score x weight) | Total.
5. Sensitivity check: would the winner change if the top-weighted criterion moved by 10 points either way?
6. Give the result, the main trade-off, and the 2-3 facts that, if they turned out different, would change the recommendation.

Stay neutral. The decision is mine; your job is to make the trade-offs clear.

Example (excerpt):
| Option | Cost (30) | Speed (20) | Quality (30) | Risk (20) | Total |
| Vendor A | 4 → 120 | 3 → 60 | 4 → 120 | 3 → 60 | 360 |
| Build in-house | 2 → 60 | 2 → 40 | 5 → 150 | 2 → 40 | 290 |

Reason for Vendor A risk = 3 (ASSUMED): no information about their uptime record.
Sensitivity: if Quality rises to 40, the gap narrows but Vendor A still wins.
What would change the answer: an in-house cost estimate below €6,000.

When to use it

Use this prompt when you are stuck between several reasonable options: choosing a software vendor, a job offer, a location, a hire or a project to prioritize. It works best when the choice involves trade-offs across several factors and you want to make your reasoning explicit, either for yourself or to explain it to others.

How to customize it

• [DECISION] should be phrased as a choice: “Which CRM should our 10-person team adopt this quarter?”
• [OPTIONS] can be two or more. More than six becomes hard to read; shortlist first.
• [CRITERIA AND WEIGHTS] is where your priorities live. Setting the weights yourself, before seeing scores, keeps the analysis honest.
• [FACTS] are what the scores are based on. The more concrete, the fewer “ASSUMED” scores you get.
• [HARD CONSTRAINTS] remove options that are simply not possible, so they do not distort the matrix.

Why it works

A weighted decision matrix is a long-established decision technique, and the prompt applies it step by step. Separating hard constraints from weighted criteria avoids a common mistake: an option that breaks the budget can still score well on everything else.

Requiring a reason for every score, and an “ASSUMED” label where the facts are missing, keeps the model from presenting guesses as analysis. The sensitivity check shows whether the result is robust or depends on one weight you set somewhat arbitrarily. Asking which facts would change the answer turns the output into a to-do list for further checking. The closing line keeps the model in an advisory role, which suits a decision that is ultimately yours.

Tips and variations

• Fill in the weights before reading the scores; changing them afterwards to get a preferred winner defeats the purpose.
• Run a pre-mortem on the winning option: see our pre-mortem prompt of the day for the method.
• For team decisions, have each person set weights separately and compare.
• Ask for the matrix as CSV so you can adjust scores in a spreadsheet.
• If the totals are within 5%, treat the options as tied and decide on the factor you care most about.