Loading page
About this tool
Compare options with weighted criteria, live scoring, rankings, sensitivity analysis, and an exportable decision report.
A decision matrix scores each option against weighted criteria and ranks them: normalise the weights so they sum to 1, multiply each rating by its weight share, and add up. This tool runs that weighted scoring model on a 1-10 rating scale, mirrors the rating for criteria where lower is better such as cost or lead time, and then does the part most templates skip — telling you which weight would have to change, and by how much, for the runner-up to win. It is for teams making a build-versus-buy, vendor or hiring call that has to be defensible afterwards.
Open Decision Matrix Tool on AltFTool — it loads instantly in your browser.
Name your criteria under Criteria and weights, set each Weight from 0 to 10, and tick Lower is better for cost or lead time.
Use Add option for up to 10 options, then rate each option from 1 to 10 against every criterion in the Ratings table.
Read the Ranking and Margin, check What would change the answer for the weight that flips the runner-up ahead, then use Copy report.
For every criterion, the exact weight at which the second-placed option would overtake the winner.
Mark a criterion lower-is-better and its rating is mirrored, so cheap scores well without inverting numbers by hand.
Scores within 0.1 are reported as effectively tied rather than dressed up as a winner.
Divide each criterion's weight by the total of all weights to get its share, multiply each option's rating by that share, then sum across criteria. With weights 5, 3 and 2 the shares are 50%, 30% and 20%, so ratings of 3, 9 and 6 give 0.5x3 + 0.3x9 + 0.2x6 = 5.4 out of 10. Because the shares sum to 1, the score stays on the same 1-10 scale as the ratings.
Rate the raw quality of that cost on the same 1-10 scale and tick lower-is-better, which mirrors it as 11 minus the rating. Never mix directions inside one matrix without flagging it — a matrix where high means good on three criteria and bad on the fourth is the most common way these get silently wrong.
Between four and seven for most decisions, and this tool caps it at 12. Past that point the weight shares get so small that no single criterion moves the result, everything scores near the middle, and the matrix stops discriminating between options rather than becoming more accurate.
Treat it as a tie and decide on something the matrix does not measure — reversibility, who has to live with it, or which option preserves more choices later. A gap under 0.1 on a 10-point scale is well inside the noise of subjective ratings, so declaring a winner there is false precision.