How to build one
- List the options you are seriously considering — three to five is plenty.
- Choose four to seven criteria. Each should be something the options really differ on.
- Give each criterion a weight from 1 to 5 for how much it matters to you.
- Score each option on each criterion from 1 to 5, using facts where they exist.
- Multiply each score by its weight and add up each option's row. The highest total is the matrix's answer.
A worked example: choosing a laptop
| Criterion (weight) | Laptop A | Laptop B | Laptop C |
|---|---|---|---|
| Price (5) | 3 → 15 | 5 → 25 | 2 → 10 |
| Battery life (4) | 5 → 20 | 3 → 12 | 4 → 16 |
| Weight (3) | 4 → 12 | 3 → 9 | 5 → 15 |
| Screen (2) | 4 → 8 | 3 → 6 | 5 → 10 |
| Total | 55 | 52 | 51 |
A wins — by three points out of seventy. That is the first thing a matrix teaches: when totals are this close, the matrix has not found a winner, it has found a tie, and a one-point change to any score would flip it.
Two ways it quietly goes wrong
- Backwards weighting. Most people already have a favourite, and adjust the weights — without noticing — until the favourite wins. If you catch yourself changing a weight after seeing the totals, you have learned what you wanted, which is useful, but the matrix did not tell you.
- Averaging away a deal-breaker. An option that scores 1 on something you cannot live without can still win on totals. Deal-breakers are not criteria; remove those options before you build the matrix.
A faster alternative
Instead of scoring everything, remove options one at a time on the thing that matters most, and look again after each removal — the elimination method. It handles deal-breakers properly and cannot average them away.
And if the slow part is finding the facts for each cell, compare⁄io does that part: type the options, say what matters, and it researches each with a live web search and fills a sourced table of at least ten attributes. Use it to fill your matrix, or skip the matrix and read the verdict.
Questions people ask
What is a weighted decision matrix?
A table that scores each option against several criteria, multiplies each score by how much that criterion matters, and adds the results. The highest total is the suggested choice.
Is a decision matrix better than a pros and cons list?
Usually, because it forces you to say how much each factor matters. A pros and cons list treats a minor pro and a major con as one item each. But a matrix can still average away a deal-breaker — remove those options first.
How many criteria should a decision matrix have?
Four to seven. Fewer and it misses things; more and the minor criteria add enough small points to drown out the ones that matter.