methodEnglish

How to complete a Prioritization Matrix

Step-by-step guidance for creating a decision-ready weighted matrix.

EnglishEspaƱolFranƧais

Documentation follows the current system language. Missing translations fall back to the page's canonical language.

Weighted Decision Matrix

How to complete a Prioritization Matrix

Use this workbench when you have several alternatives and need a transparent, evidence-aware way to rank them using weighted criteria.

Frame the decision.
Define comparable alternatives.
Define distinct criteria.
Set and confirm weights.
Score each alternative.
Add rationale, evidence, and confidence.
Review ranking and disagreement.
Run sensitivity scenarios.
Generate the decision report.

Decision setup

Frame one decision

A good matrix answers one clear decision question. State the owner, stakeholders, constraints, assumptions, and what a good decision means.

Required fields

  • Decision question
  • Context or objective
  • Owner or group
  • Constraints
  • Assumptions

Quality check

  • If the question combines multiple choices, split it before scoring.

Alternatives

Compare options that can be selected

Each alternative should be clear enough to score against the same criteria. Include status quo when it is a real option.

Alternative fields

  • Name
  • Description
  • Category
  • Owner
  • Status
  • Evidence notes

Review rule

  • If an alternative is actually a criterion, move it to the criteria step.

Criteria and weights

Define what matters and how much

Criteria should be distinct, measurable, and not biased toward one option. Weights should sum to 100 percent; Albarena normalizes transparently when they do not.

Recommended criteria count

  • 4 to 8

Warnings

  • Duplicate criteria
  • Vague criteria
  • Dominant weights
  • Unintentional equal weights

Scoring

Score higher as better by default

Use the 1 to 5 scale unless there is a reason for 0 to 10. Cost, risk, complexity, and effort should usually be phrased as attractiveness or reduction so 5 still means better.

Cell fields

  • Score
  • Rationale
  • Confidence
  • Evidence
  • Participant

Review rule

  • AI-drafted scores remain draft until confirmed.

Results and sensitivity

Do not overstate precision

The ranking is a decision aid, not a mechanical verdict. Close results and sensitivity changes should trigger discussion.

Close call

  • under 5 percent difference

Sensitivity checks

  • Equal weight
  • Cost focused
  • Impact focused
  • Risk averse
  • Speed focused

Report

Explain the tradeoff

The report should show the ranking, why the top option wins, important tradeoffs, uncertainty, disagreement, evidence gaps, risks, assumptions, and next steps.

Report tone

  • Concise
  • Defensible
  • Clear about uncertainty