Make a hard call with a weighted decision matrix
FreeA ChatGPT prompt that builds a weighted decision matrix — scoring your options against weighted criteria — and returns a ranked, defensible recommendation plus a sensitivity check on what would flip the call.
Turns a hard, multi-option choice into a transparent weighted decision matrix with a ranked, defensible recommendation and a check on what would change the outcome.
This prompt
You are a decision analyst who helps people make high-stakes choices with weighted decision matrices. You stay neutral, show your math, and would rather flag a missing fact than invent a score.
Your task: build a weighted decision matrix for the decision below and deliver a recommendation the reader can defend. You have succeeded when the reader can see exactly why the winner won, how close the runner-up was, and what single change would flip the call.
Here is the decision to analyze:
<decision>
Decision and desired outcome: {{decision}}
Options under consideration: {{options}}
Criteria that matter (may be blank): {{criteria}}
Deal-breakers / non-negotiables (may be blank): {{deal_breakers}}
Context — budget, timeline, who is affected, top priorities: {{context}}
Scoring scale: {{scoring_scale}}
</decision>
Work through these steps:
1. Restate the decision in one line and define what a good outcome looks like. If the goal is ambiguous, state the interpretation you will use.
2. Screen for deal-breakers first. List each non-negotiable and mark every option pass or eliminated. An eliminated option leaves the matrix (note why); only survivors get scored, no matter how well they would have scored elsewhere.
3. Settle on 4–8 evaluation criteria that are distinct and do not overlap. Refine any supplied criteria; propose sensible ones drawn from the decision if none were given. Phrase each so that higher is always better — for cost, risk, or effort, score so the cheaper, safer, or easier option earns the higher number.
4. Assign each criterion a weight so the weights total 100. Let the stated priorities drive the spread and force real differences — avoid giving everything a similar weight. Add a one-line reason for each weight.
5. Score every surviving option on every criterion using the {{scoring_scale}} scale, with a short anchor for what a low versus high score means. Score each cell on its own evidence before looking at any totals.
6. For each cell, multiply the score by the weight, then sum each option's weighted values into a total. Show the totals on a comparable scale.
7. Rank the options. Then run a sensitivity check: name the smallest weight or score change that would flip first place, and how large it would need to be. Flag any top-two gap under about 5% as a near-tie.
8. Recommend one option and say what it depends on. Name the runner-up, the condition under which it would win instead, and the strongest reason to reconsider.
Present the result in this order:
- Decision & success definition
- Deal-breaker screen — each non-negotiable, each option marked pass or eliminated
- Criteria & weights — a table: criterion | weight | why it is weighted that way
- Weighted matrix — a table with one row per surviving option and, for each criterion, the raw score and the weighted value, ending in a Total column
- Ranking — options ordered from best to worst total
- What would flip it — the sensitivity finding
- Recommendation — the pick, what it hinges on, the runner-up and its winning condition, and watch-outs
- Assumptions & missing info — facts you had to assume, and questions that would sharpen the scoring
Constraints:
- Score each cell before summing; never work backward from a preferred winner to justify its scores.
- Keep the weights totalling 100 and every score inside the {{scoring_scale}} range.
- Everything that moves the ranking must appear in the table — no hidden criteria.
- When you lack a fact needed to score a cell, mark it as an assumption or ask for it, rather than presenting a made-up number as certain.
- Write in plain language and define any term a smart non-specialist would not know.8 fields still empty — the {{like_this}} slots fill in as you type.
See it in action
What this prompt produces with example details filled in:
**Decision & success definition** Pick one CRM for the 40-person sales team, live within 60 days, that reps use daily, gives leaders reliable forecasting, and scales toward ~80 seats within budget. I'm reading "good outcome" as, in order: daily rep adoption first, trustworthy reporting second, headroom to grow — all under about $120/user/month. **Deal-breaker screen** | Non-negotiable | HubSpot | Salesforce | Pipedrive | |---|---|---|---| | Native two-way Marketo sync | Pass | Pass | Pass | | Live within 60 days | Pass | Pass (phased rollout) | Pass | | SOC 2 Type II | Pass | Pass | Pass | All three survive and get scored. Note: if any had failed a row here, it would leave the matrix regardless of how well it scored elsewhere — that's the point of screening first. **Criteria & weights** (proposed from the decision, since none were supplied) | Criterion | Weight | Why | |---|---|---| | Ease of adoption (daily rep use) | 25 | Adoption failed last time; non-technical reps — the top stated priority | | Reporting & forecasting | 20 | Data-driven leadership needs trustworthy pipeline numbers | | Integrations with our stack | 20 | Must fit Marketo and existing tools to be usable on day one | | Total cost of ownership | 15 | Real budget cap; licenses + admin + add-ons | | Scalability to ~80 seats | 12 | Headcount doubling within 2 years | | Vendor support & reliability | 8 | Matters, but least differentiating among these three | | **Total** | **100** | | **Weighted matrix** (scores 1–10, higher = better; cost scored so cheaper = higher. Weighted value = score × weight ÷ 100) | Option | Adoption (25) | Reporting (20) | Integrations (20) | TCO (15) | Scalability (12) | Support (8) | Total | |---|---|---|---|---|---|---|---| | HubSpot | 9 → 2.25 | 8 → 1.60 | 8 → 1.60 | 6 → 0.90 | 8 → 0.96 | 8 → 0.64 | **7.95** | | Salesforce | 5 → 1.25 | 10 → 2.00 | 9 → 1.80 | 4 → 0.60 | 10 → 1.20 | 7 → 0.56 | **7.41** | | Pipedrive | 9 → 2.25 | 6 → 1.20 | 6 → 1.20 | 9 → 1.35 | 6 → 0.72 | 7 → 0.56 | **7.28** | **Ranking** 1. HubSpot — 7.95 2. Salesforce — 7.41 3. Pipedrive — 7.28 **What would flip it** HubSpot leads Salesforce by 0.54 (about 7%) — not a near-tie, but not comfortable either. The gap is driven almost entirely by ease of adoption (9 vs 5). Move 10 points of weight from Adoption (25→15) to Reporting (20→30) — i.e., decide analytics matter more than self-serve usability — and Salesforce wins, 7.91 to 7.85. So the real question isn't "which CRM is best," it's "do we value daily rep adoption more than deep forecasting?" Pipedrive only takes the lead if Total cost of ownership is weighted above ~40%, far beyond its stated importance. **Recommendation** Go with **HubSpot Sales Hub**, as long as ease of adoption stays your top priority — it wins on the exact factor your last rollout died on, while staying competitive on reporting and integrations. Runner-up: **Salesforce**, which becomes the pick if leadership decides forecasting depth and scale to 80+ outweigh day-one usability (see the flip above). Pipedrive is the value play but gives up too much reporting and integration depth for a data-driven team. Watch-outs: the TCO scores are estimates — get firm per-seat quotes (including required add-ons) before signing, since a price surprise is the most likely thing to move this ranking; and pressure-test Salesforce's 60-day go-live, which I scored as a phased rollout, not a full one. **Assumptions & missing info** - TCO scores assume near-list pricing; real quotes could swing HubSpot vs Pipedrive on cost. - Assumed all three currently hold SOC 2 Type II — confirm each vendor's report. - No weight was given to mobile experience or specific report types; if those matter to you, add them as criteria and re-run.
Tips
- List your true deal-breakers in the deal_breakers slot — they knock options out before scoring, so a cheap-but-unqualified option can't sneak a win on points.
- Leave the criteria slot blank the first time to see which factors the model surfaces, then re-run with your own refined list.
- Read the 'what would flip it' section closely — if a tiny weight change swaps the winner, the top two are effectively tied, so pick the one that's cheaper to reverse.
- Re-run with different weightings (cost-first vs. quality-first) to see how sensitive the winner is to your priorities.
Built by Web Innoventix
Want the work done, not just prompted? We design, build and rank websites that get found on Google and cited by AI.
Get a free quoteMore prompts
Browse all →Pressure-test your business idea before you build
Stress-test any business idea across demand, moat, unit economics, its riskiest assumption, and the cheapest test to run this week — and get an honest build/refine/kill verdict, not a pep talk.
Nail your positioning statement and value proposition
A paste-and-go ChatGPT prompt that interviews you, then writes a specific, differentiated positioning statement and value proposition your best-fit buyer instantly recognizes as built for them.
Design a pricing page: tiers, plan names, and anchoring
A ChatGPT prompt for a pricing page that interviews you about your product, then designs the whole thing — the value metric, three to four named and charm-priced tiers, the "most popular" anchoring, CTA copy, and what to test first.
Write a one-page executive summary from a long report
The prompt to write an executive summary from a long report: it produces a one-page brief with the bottom line up front, the numbers that matter, and prioritized next steps for your specific reader.
Turn messy meeting notes into decisions and action items
Paste raw, disorganized meeting notes and get back a decision-ready record where every decision is stated plainly and every action item has an owner and a due date.
Turn a feature idea into a developer-ready PRD
A ChatGPT prompt that writes a PRD: it interviews you about your feature idea one question at a time, then produces a developer-ready product requirements document with user stories, testable requirements, and success metrics.

