Overview
A decision matrix turns a fuzzy comparison into an explicit one. You list the options as rows, the criteria that matter as columns, weight each criterion by how much it counts, score every option on every criterion, and multiply through. The winner is not chosen by whoever argues loudest but by the arithmetic of what you said you valued.
Its real benefit is not the number at the bottom — it is the argument the matrix forces you to have about weights. Deciding that price matters twice as much as speed, before you see the scores, prevents you from quietly rigging the criteria to justify the option you already liked.
When to use it
You have several viable options and multiple criteria that matter differently — tools, vendors, job offers, features.
How to use it
List the options
The realistic choices, as rows.
List the criteria
The factors that genuinely matter, as columns — keep it to the vital few.
Weight the criteria
Assign each a weight (say 1–5) reflecting importance. Do this before scoring.
Score each option
Rate every option on every criterion on a fixed scale.
Multiply and total
Multiply score by weight, sum each row, and read the ranking — then sanity-check it against your gut.
Worked example
Choosing a CRM: criteria are price (weight 5), ease of use (4), integrations (3), support (2). Three vendors are scored 1–5 on each. Vendor B loses on price but wins on ease of use and integrations; weighted, it edges ahead. Crucially, the team set the weights first, so nobody could inflate ‘integrations’ afterwards to rescue a favourite.
Common pitfalls
- Setting weights after seeing scores, which lets bias back in.
- Too many criteria, which dilutes the ones that truly decide.
- Treating the total as gospel — if the winner feels wrong, interrogate your weights, don’t override silently.
Frequently asked questions
What scale should I use?
Any consistent one — 1–5 is common. Consistency matters more than the range.
What if two options tie?
A tie usually means a criterion is missing. Add the factor that actually separates them in your mind.
Isn’t this just false precision?
The numbers are rough by design; the value is the structured conversation about weights, not decimal accuracy.