The structure of a decision matrix
A working decision matrix has four moving parts:
- Options (rows). The choices you're deciding among. Houses, candidates, vendors, vacation destinations.
- Criteria (columns). The factors that matter to you. Price, commute, schools, condition, maintenance, etc.
- Weights. A multiplier for each criterion (0–10 or 0–1) reflecting how much it matters to you. The whole point of the matrix is making weights explicit.
- Scores. A rating (typically 1–10) of how each option performs on each criterion.
The total score for each option is the sum of (score × weight) across every criterion. The matrix ranks options by total score. You read them top to bottom and you have your answer — or, more usefully, you have the math behind your gut.
What it's actually for
Decision matrices aren't truly about computing the right answer. They're about surfacing the trade-offs. Three things happen almost every time someone runs one honestly:
- You realize you've been weighing a criterion (commute, square footage, school district) much more heavily than you'd admit out loud.
- An option you'd half-dismissed scores surprisingly well. Worth a second look.
- Your "favorite" option scores lower than expected — and the matrix tells you which criterion is dragging it down.
None of those require you to obey the spreadsheet. They just put the choice on the table where two people can argue about it instead of talking past each other.
When to use one
- Major purchases. Houses, cars, mattresses, appliances — anywhere there are more than three serious options.
- Career decisions. Competing job offers, especially when comp is only one factor.
- Vendor selection. Software, contractors, agencies.
- Family decisions. Schools, vacation destinations, neighborhoods — anywhere multiple people have different weights and need to negotiate.
- Hiring. Most structured interview scoring is a decision matrix in disguise.
How to build one in twenty minutes
- List your options. Three to seven is the sweet spot; ten gets unwieldy.
- Brainstorm criteria — what actually matters about this choice. Aim for 5–10. More than that and the weights stop mattering.
- Assign weights. The instinct to give everything weight 5 is the mistake the matrix exists to fix. Make some things 9 and some things 2.
- Score each option on each criterion. Score in isolation, not relative to other options — "this house's commute is a 6," not "this house's commute is better than that one."
- Read the totals. Look at the top two or three. Then look at where they differ and ask if those differences match your gut. If they don't, the weights are wrong — fix them and re-score.
Pitfalls
- Reverse-engineering the answer. If you already know which option you want and you adjust weights until it wins, you've built a justification, not a decision aid.
- Too many criteria. Past about ten, every option's strengths and weaknesses average out and the totals flatten. Trim to what really matters.
- Vague criteria. "Better fit" isn't scorable. "Within 30 minutes of work" is.
- Skipping weights. An unweighted matrix treats "school district" and "kitchen cabinet color" as equally important. They are not.
- Forgetting the deal-breakers. Some criteria are thresholds, not weights — "must be under $450K" knocks options out regardless of their score elsewhere. Apply those first.
Variants worth knowing
- Pugh matrix. Scores everything relative to a baseline option using only +, 0, and −. Faster, less precise.
- Decision tree. For sequential choices where each decision changes the next set of options.
- Pros/cons list. The pre-spreadsheet ancestor. Useful when you only need to convince yourself, not a partner or a spouse.
Related templates and concepts
Decision matrices show up everywhere — see the dedicated decision helpers for the choices people most often run them on, plus generic matrices for one-off decisions. The templates for homebuyers hub and templates for college students hub both make heavy use of decision matrices.
