Decision Matrix for Tough Calls: Strip the Emotion Out of the Comparison
Tough decisions rarely stall because you lack options. They stall because you can't hold every option in your head at once and compare them consistently. Three job offers, two suppliers, one partnership decision, and the criteria you're weighing keeps shifting: price wins one day, trust wins the next, and nothing ever pulls ahead for good. The real problem isn't indecision, it's comparing options before you've locked down how much each criterion actually matters. The AI doesn't make the call here; it takes the criteria you already care about, fixes their weight, and scores every option against the same ruler. Then it reverse-tests your final pick: 'what's the scenario where you can't defend this decision.' This prompt won't decide for you. It quiets the noise in your head and makes your actual preference visible.
You are an impartial decision analyst and strategic advisor. Your job isn't to make the call for the user; it's to compare their options in a transparent matrix built from criteria they define themselves, surface any emotional bias hiding in the inputs, and stress-test the final pick with a reverse scenario. You're not handing over "the right answer." You're showing the user the actual ground their preference stands on. Work from the following information: - The options being compared (2 to 5): [briefly describe each one] - The criteria that matter to you: [e.g. price, risk, time, trust, long-term impact; at least 3 criteria] - How you'd rank each criterion by importance, if you already know: [most important to least, leave blank if unclear] - Your decision deadline: [date or time window] RETURN THE OUTPUT IN EXACTLY THIS ORDER, WITH THESE EXACT HEADERS: === CRITERIA WEIGHTING === Weight the criteria given (percentages or a 1-5 scale, pick one consistent system). Briefly explain the logic behind each weight. Stay true to the priority order the user gave; don't impose your own ranking. === COMPARISON MATRIX === Score every option against every criterion (e.g. 1 to 10), with a short justification for each. Table or clean bullet format, one sentence max per cell. === WEIGHTED RESULT === Rank the options by weighted total score. If the gap is small (under 5%, say), state that plainly. Don't be afraid to say "there's no measurable difference between these two." === EMOTIONAL BIAS CHECK === Flag any sign in the user's inputs of emotional attachment to one option (habit, familiarity, fear of loss). If there's nothing, write "no clear bias signal detected." Don't manufacture one. === REVERSE TEST === Assume the user picks the highest-scoring option. Answer this with a concrete example: "what's the scenario where you can't defend this decision three months from now." Make it specific enough to actually make the user think. === DECISION LOG ENTRY === Write a 3-4 line summary the user can copy straight into their own decision log: the call made, the biggest known risk at the time, and when it should be revisited. RULES: - No made-up statistics, no fake success rates, no unverifiable claims like "X% of people who used this method made the right call." - If the user didn't provide criteria, don't invent your own. Ask what they actually care about first. - Don't secretly favor any option in the matrix; every score needs a consistent, stated logic behind it. - Keep it tight. Skip the philosophical preamble and get straight into the analysis.
How to use it
- List the options you're comparing (2 to 5) and the criteria you actually care about (price, risk, time, trust, whatever applies), in plain terms. If you don't set the criteria yourself, the AI will invent generic ones and the whole matrix gets weaker.
- Run the prompt and look at how each criterion gets weighted and scored. Be honest with yourself: if a weight feels off, change it and rerun. The goal isn't a result that feels 'right,' it's one that feels real.
- Don't skip the reverse test. If there's a scenario within a week where you couldn't defend the outcome, write it in your decision log and put a review date on your calendar.
Example / tip
A small business owner is torn between two software vendors: one is cheap but slow on support, the other is pricier but integrates fast. She enters price, support speed, integration time, and long-term lock-in risk as her criteria. The matrix scores each vendor against all four, and in the weighted total the pricier vendor edges ahead slightly, because integration time got flagged as high-weight. The reverse test asks: 'If you pick the cheap vendor and lose a customer over a support delay, can you defend that decision three months from now?' That question surfaces what she actually cared about: not price, but risk tolerance.
Usage example
Copy the prompt as is, fill in the bracketed fields for your own business, and paste it into ChatGPT or Claude. The more concrete your input, the more useful the output. Take the first draft, ask for a one-sentence fix on whatever you don't like, then give it a final pass in your own voice.
When not to use it
Don't use this as your only decision tool for irreversible life-and-health calls (a medical treatment choice, a marriage or divorce). The criteria matrix is built for business and operational decisions; for deeply personal ones, professional guidance and your own judgment should carry more weight.
Output quality checklist
- Do the criteria weights actually reflect what you care about, or did the AI just default to its own ranking?
- Are the scoring justifications concrete, or do they read like vague 'this one just seems better'?
- Did the reverse-test question surface a scenario that genuinely unsettles you, or did it stay surface-level?
