Problem solving / RCA
Weighted decision matrix
Weighted decision matrix is licensed CC BY 4.0. Attribution: Katafacts (katafacts.com).
Customise with AISkill file ↓1 · What it is
What it is
A weighted decision matrix scores a small set of options against criteria that don't all matter equally — cost, ease of adoption, support quality, whatever the real decision actually turns on — each criterion carrying its own weight, each option scored directly against each criterion (1 to 5). The weights and scores are multiplied and summed into a single composite score (0 to 100), which sorts options into priority, consider, or deprioritize tiers. It's genuinely different from a countermeasure matrix's quick 2-axis impact/effort filter — this is for a real, deliberated choice between a handful of options, not a fast triage pass — and from an FMEA, which ranks failure modes by risk, not options by fit.
2 · When to use it
When to use it — and when not to
Use it when
- You have a real handful of options (not two obvious extremes) and criteria that genuinely matter differently to the decision, not equally.
- You can state real weights and score each option honestly, even when the highest-weighted criterion doesn't have an obvious winner.
- You want the ranking to survive someone asking 'why did we pick this one' with a specific, defensible answer.
Not when
- You're ranking failure modes, not comparing options against each other — that's an FMEA's job, not this one's.
- You need a fast impact/effort triage across many small ideas, not a deliberated choice between a few real options — a countermeasure matrix's simpler 2-axis filter fits that better.
- The decision is really fixed by a hard constraint (budget, policy, a non-negotiable requirement) — state the constraint and eliminate options first; this tool compares genuinely viable options, not options one of which was never really in play.
3 · How to fill it in
How to fill it in
- Decision
- What's being decided, and what are you scoring options against?
- Criteria and weights
- What matters in this decision, and how much does each criterion count? Weights should sum to 100% — your own judgement about what matters, never AI-assigned.
- Options, scored
- For each option, score it 1 (weakest) to 5 (strongest) against every criterion — your own honest judgement, never AI-scored. The composite score and tier are computed from your weights and scores, never entered directly.
- Narrative
- Drafted from the computed ranking, naming what's actually driving the top option's score.
4 · What good looks like
What good looks like
The example below scores three candidate CI-software vendors against four weighted criteria — and deliberately shows an option that wins the single most heavily weighted criterion outright but still doesn't clear the priority threshold overall, because the composite score is what actually decides, not any one axis.
Same example, as a downloadable xlsx workbook.
Download .xlsxWeighted decision matrix
CI software vendor selection
Selecting a continuous-improvement software platform to replace ad hoc spreadsheet tracking for kaizen events and A3s across the plant floor.
Elena Cho, Process Engineer · 2026-03-18
Decision
Selecting a continuous-improvement software platform to replace ad hoc spreadsheet tracking for kaizen events and A3s across the plant floor.
Criteria and weights
- Cost35%
- Ease of adoption20%
- Support quality20%
- Integration fit25%
Options, scored
GembaFlow
75/100 · PriorityCost: 2/5 · Ease of adoption: 5/5 · Support quality: 4/5 · Integration fit: 5/5
LeanTrack
74/100 · ConsiderCost: 5/5 · Ease of adoption: 3/5 · Support quality: 3/5 · Integration fit: 3/5
Kaizen360
72/100 · ConsiderCost: 3/5 · Ease of adoption: 4/5 · Support quality: 5/5 · Integration fit: 3/5
Narrative
GembaFlow is the vital-few pick here, at 75/100 — the only option that clears the priority threshold, driven by the strongest ease-of-adoption and integration-fit scores in the field, the two highest-weighted criteria after cost. LeanTrack is the most tempting single-axis pick — it wins cost outright, the most heavily weighted criterion — but that strength alone isn't enough to carry it: weaker adoption, support, and integration scores pull its composite down to 74, just under the priority line. Worth a second look only if cost turns out to matter more in practice than this weighting reflects. Kaizen360 trails both, held back mainly by a weaker integration-fit score despite the strongest support-quality rating of the three.
5 · Common mistakes
Common mistakes
Weights don't sum to 100% and nobody checks.
A set of weights that doesn't sum to 100% silently distorts every composite score in a way that's hard to spot just by looking at the final ranking — the check is cheap and the failure mode isn't.
Scoring options relative to each other from a gut feel about which one 'seems best,' instead of scoring each one independently against each criterion.
A relative gut-feel ranking just launders the same intuition the tool exists to check — real per-criterion scores are what let the weighting actually change the outcome instead of rubber-stamping a decision already made.
Letting the option that wins the single highest-weighted criterion decide the outcome without checking the actual composite.
The whole point of weighting several criteria is that the strongest option on one axis can still lose on the composite — treating one criterion as decisive defeats the purpose of scoring the others at all.
6 · What it connects to
What it connects to
upstream
Countermeasure matrix
When a countermeasure matrix's fast impact/effort filter narrows a long list down to a genuine handful of finalists, a weighted decision matrix is where those finalists get compared in depth on the criteria that actually matter.
FMEA
When an FMEA's recommended actions produce more than one real candidate fix for the same risk, a weighted decision matrix is where you choose between them on more than just risk priority number alone.
downstream
Business case
Once a weighted decision matrix names the option worth pursuing, a business case is where that choice gets justified with real investment and value numbers.
7 · Where AI helps
Where AI helps
Judgement — stays yours
- Deciding the real weight each criterion deserves
- Deciding an honest score for each option against each criterion
Analysis — AI helps
- Drafting the narrative from the ranking actually computed, naming what's driving the top option's score
Drudgery — automated
- Computing each option's composite score from its weights and scores, every time one changes
- Sorting options into priority, consider, and deprioritize tiers
- Exporting to xlsx in the house format
9 · Rate this kata