Decision Matrix: Template + Worked Example

What is a decision matrix?

A decision matrix is a grid: options in the rows, criteria in the columns, a weight on each criterion and a score in each cell. Multiply, add across, and you get one number per option that reflects your own stated priorities.

The value is not the final number. It is that the matrix forces three things most decisions skip:

• Naming the criteria. You have to say what you actually care about before you look at the options.
• Ranking the criteria. Weights make you admit that salary matters more than commute, or less.
• Comparing like with like. Every option is judged on the same dimensions, so the shiny one can’t win on charm alone.

If you’d rather fill one in than read about it, the interactive decision matrix does the math for you. The rest of this page explains how to build one well and when not to trust it.

When should you use a decision matrix?

Use a decision matrix when you have three or more real options and several criteria that pull in different directions. Two options with one clear trade-off don’t need a table; twenty options need a filter first.

Good fits:

• Choosing between job offers, apartments, schools or cities.
• Picking a vendor, a tool or a hire from a shortlist.
• Prioritizing projects when budget or time is fixed.

Before you build one, check how much process the decision deserves. The Decision Weight Score rates Reversibility, Impact and Uncertainty on a 1 to 3 scale and turns them into a 1 to 10 score. A DWS of 1 to 3 means decide in under 10 minutes and skip the matrix. A 4 to 6 is the sweet spot for a quick matrix. A 7 to 10 deserves a matrix plus a pre-mortem and a written journal entry.

How to build a decision matrix (step by step)

Building a weighted decision matrix takes six steps and, for most decisions, under an hour.

1. List the options. Include the ones not on the table yet: “stay where I am”, “wait six months”, “negotiate”. The default option is often the strongest competitor and the easiest to forget.
2. List the criteria. Ask what a good outcome looks like a year from now. Write each criterion so that a higher score is always better (“low cost”, not “cost”).
3. Weight the criteria before you score. Split 100 points across them. Do this before you look closely at the options, or you will tune the weights to fit the option you already like.
4. Score each option on each criterion. Use a 1 to 5 scale. Score one criterion at a time across all options (a column at a time), not one option at a time. It keeps the comparisons honest.
5. Multiply and add. Weight times score for each cell, then sum each row. Divide by 100 to get a weighted average on the same 1 to 5 scale.
6. Stress-test the winner. Change the top two weights by 5 to 10 points and see if the winner changes. If it does, the decision is close, and the real question is which of those two criteria matters more.

Decision matrix template

Here is the blank template. Copy it into a spreadsheet, or use the decision matrix tool.

Rules that keep it clean:

• Weights sum to 100.
• Scores run 1 (poor) to 5 (excellent).
• Deal-breakers are not criteria. If an option fails a must-have, remove it before scoring.

Worked example: choosing between three job offers

Say you have three job offers. This is a hypothetical example, with numbers chosen to show how the method works.

• Option A: a large, established company. Good pay, slow growth, office five days a week.
• Option B: an early-stage startup. Lower base pay, fast learning, a hybrid schedule, real risk.
• Option C: a mid-size company, fully remote. Good pay, decent growth, a team you liked in interviews.

You decide on five criteria and weight them before scoring:

How the totals come out:

• A: (30 × 4) + (25 × 2) + (20 × 3) + (15 × 2) + (10 × 5) = 120 + 50 + 60 + 30 + 50 = 310.
• B: (30 × 3) + (25 × 5) + (20 × 4) + (15 × 3) + (10 × 2) = 90 + 125 + 80 + 45 + 20 = 360.
• C: (30 × 4) + (25 × 3) + (20 × 4) + (15 × 5) + (10 × 3) = 120 + 75 + 80 + 75 + 30 = 380.

Option C wins, but not by much. So run the stress-test from step 6.

The sensitivity check

Suppose you move 10 points from flexibility to growth: growth becomes 35, flexibility becomes 5.

Now the startup wins. That is the real finding of this matrix: the choice between B and C comes down to one question. How much do you value learning speed over flexibility at this point in your life? Option A loses under both weightings, so you can drop it with confidence.

This is what a good decision matrix does. It rarely hands you an answer. It shrinks a five-way tangle into one clear question you can actually think about.

When does a decision matrix fail?

A decision matrix fails when its inputs are worse than your judgment: bad criteria, rigged weights, or numbers that pretend to be more precise than they are. Here are the common failure modes.

Weights chosen after the fact

If you score the options first and then set the weights, you will, without noticing, pick weights that crown your favorite. This is confirmation bias with arithmetic on top. Set weights first and write them down.

Double counting

“Salary”, “bonus” and “total compensation” are one criterion wearing three hats. Overlapping criteria quietly multiply the weight of one factor. Merge anything that would move together.

False precision

A 3.62 against a 3.58 is a tie. Scores of 1 to 5 are rough judgments, and the error in each one is bigger than that gap. Treat anything within about 0.2 on a 1 to 5 scale as a coin flip and decide on other grounds.

Deal-breakers treated as weights

A matrix is compensatory: a great score in one place can make up for a terrible score in another. That is wrong for must-haves. If a job requires moving to a city your partner won’t move to, no salary score fixes that. Filter deal-breakers out first.

Missing options

The matrix only ranks what you put in it. If the best choice was “negotiate a remote version of offer B”, no amount of scoring will surface it. Spend real time on step 1, and ask what the opportunity cost of each option is: what you give up by choosing it.

Scores anchored by the first option

Whichever option you score first sets the reference point for the rest. This is anchoring bias. Scoring column by column (all options on one criterion) instead of row by row reduces it.

Hard-to-score factors

Some things don’t fit on a 1 to 5 scale: a gut sense about a founder, a family situation, an ethical line. Don’t force them in and pretend they’re handled. Note them beside the matrix and weigh them openly at the end.

Decision matrix vs other decision tools

A decision matrix is best for comparing several options on known criteria. Other tools are better when the problem is different.

They combine well. Use the matrix to pick, the pre-mortem to stress-test the pick, and a decision journal to record why.

How to make the result stick

Write down the matrix, the winner, and one sentence on what you expect to happen and how confident you are. Then set a date to look back. A matrix you can revisit is a learning tool; a matrix you throw away is just a calculation.

If you’d like a guided version, the Mentarium Decision Coach walks you through building the criteria and weights inside Claude or ChatGPT, and the full method lives in our guide to the decision-making process.

The Mentarium take

The decision matrix is overrated as a calculator and underrated as an interrogation. The number at the bottom matters less than the moment you have to write “Growth: 25” and realize you’d actually trade a lot of salary for it. That moment of admission is the whole point.

In the Mentarium Decision Loop (Frame → Map → Stress-test → Decide → Review), the matrix belongs in Map. It lays out options and what each costs. It does not replace Stress-test. Once the matrix names a winner, run the Bias Radar against it: for a career choice, check status quo bias, loss aversion, sunk cost fallacy, anchoring bias, bandwagon effect, overconfidence bias and the availability heuristic. A matrix built by a biased mind returns a biased answer with two decimal places.

Our rule of thumb: if the sensitivity check flips the winner, stop scoring and start thinking about the one criterion that flipped it. If it doesn’t flip, decide, write it down, and stop reopening the question.