In short: A weighted decision matrix multiplies each option's score on each criterion by that criterion's weight and adds the results; the option with the highest weighted total comes first. In the example, Austin scores 54 of a possible 70 (77.1%), only 1.4 points ahead of Denver, which is close enough to check the weights before deciding.
How to use the decision matrix calculator
This is how to make a decision matrix with it, in four steps:
- Name the options down the left: the cities, vendors, tools or candidates you are choosing between. Add up to 12.
- Name the criteria across the top, the things that decide it, and give each a weight. Any numbers work: 1 to 5 for importance, or percentages. The calculator shows each weight's share of the total, so you can see that cost carries 35.7% of the decision in the example.
- Score every option on every criterion on the scale you chose. Higher always means better. An option with a blank or out-of-range score is left out of the ranking until it is fixed, and the note says which.
- Read the ranking and the close call. The table shows each option's weighted points per criterion, its total and its share of the most possible. Under it, the calculator works out the weight on each criterion at which the runner-up would draw level. Copy it as text or CSV, or share a link that opens the same matrix.
Everything updates as you type, and nothing leaves your browser: the share link keeps the matrix in the part of the address after the #, which browsers never send to a server. It replaces a decision matrix template: the arithmetic and the ranking are done for you, and the close-call check is one a spreadsheet does not do unless you build it.
The weighted decision matrix formula, with an example
For each option, multiply its score on each criterion by the criterion's weight and add them up. A team choosing between 3 cities for an offsite, each scored 1 to 5 on cost (weight 5), travel time (4), venues (3) and weather (2), gives this decision matrix example:
| Rank | Option | Cost (5) | Travel time (4) | Venues (3) | Weather (2) | Weighted total | % of the most possible |
|---|---|---|---|---|---|---|---|
| 1 | Austin | 20 | 12 | 12 | 10 | 54 | 77.1% |
| 2 | Denver | 15 | 16 | 12 | 10 | 53 | 75.7% |
| 3 | Chicago | 20 | 20 | 9 | 2 | 51 | 72.9% |
Austin scored 4 on cost, so that criterion gives it 4 x 5 = 20 points; with travel time 12, venues 12 and weather 10, its total is 54. The most any option could score is 5 on everything, 5 x 14 = 70, so 54 is 77.1%. Chicago has the best travel time and ties for the best cost, but its weather score of 1 pulls it down to 51.
The weights do real work. With every weight set to 1, the same scores put Denver on 16, Austin on 16, Chicago on 13, a tie at the top. Weighting is how a team says out loud that cost matters more than weather, and the matrix then holds everyone to it.
Close calls: what would change the result
Austin leads Denver by 1 point, 1.4 percentage points of the most possible. The calculator finds, criterion by criterion, the weight that would close that gap with every other weight unchanged. Austin scores 4 on cost and Denver 3, so lowering the cost weight from 5 to 4 ties them; Denver is ahead on travel time, so raising that weight from 4 to 5 does the same. Both score the same on venues and weather, so those weights cannot separate them at all.
A change of 1 on a weight someone picked in a minute is not much, and that is the point. NASA's systems engineering handbook, in its section on decision analysis, asks for an assessment of the robustness of the ranking with every recommendation, and says that when the alternatives do not significantly differ, the recommendation should include all the closely ranked ones. The same handbook warns that a lower or close score sometimes deserves to win, and that when it does, the criteria should be looked at again: not only the weights, but what each one measures.
Choosing criteria, weights and a scale
Three to seven criteria are usually enough; more and the small ones add noise. Each should be something the options really differ on, and none should count the same thing twice (price and total cost of ownership, say). Agree the weights before anyone scores, so nobody tunes them to a favorite. Then write down what each score means on each criterion: for travel time, 5 might be under two hours door to door and 1 over six. A 1 to 5 scale is quick; 1 to 10 or 0 to 10 gives room to separate options that are close.
A decision matrix is for choosing between options on several criteria at once. It is not the Eisenhower matrix, which sorts tasks by urgent and important, and it is not a vote: if a group disagrees on the options themselves, each person's ranking can go into the ranked choice voting calculator instead. To keep the meeting where you fill it in on schedule, use the meeting timer, and the meeting cost calculator shows what a long debate over the weights costs.
A decision matrix in Excel or Google Sheets
In a spreadsheet the weighted total is one function. With the weights in B2:E2 and an option's scores in B3:E3, =SUMPRODUCT($B$2:$E$2,B3:E3) gives its total; fill it down for the other options. Excel's SUMPRODUCT returns the sum of the products of corresponding ranges, and Google Sheets' SUMPRODUCT works the same way. The calculator's Copy as CSV pastes the weights, the weighted points, the totals and the ranks into either one.
Esta calculadora también está en español: matriz de decisión ponderada.
Questions people ask
What is a weighted decision matrix?
A table with the options down the side and the criteria across the top. Each criterion gets a weight for how much it matters, each option gets a score on each criterion, and an option's weighted total is the sum of its scores times the weights. The option with the highest total ranks first. It is also called a weighted scoring model, a prioritization matrix or a decision grid.
What is the decision matrix formula?
Weighted total = score on criterion 1 x weight 1 + score on criterion 2 x weight 2, and so on for every criterion. The calculator also shows each total as a percentage of the most possible, which is the top of the scale times the sum of the weights: 70 in the example, so Austin's 54 is 77.1%. Weights can be any numbers: 5, 4, 3 and 2 give the same ranking as 50, 40, 30 and 20, or as percentages that add up to 100.
How close is too close to call?
The calculator flags a close call when the top two are within 5 percentage points of the most possible (you can change the margin). That is a prompt to look harder, not a rule: the scores are judgments, and a one-point change on one criterion can be worth more than the gap. The list under the ranking shows, for each criterion, the weight at which the runner-up would draw level, so you can see whether the result rests on a weight nobody is sure of.
How do I make a decision matrix in Excel or Google Sheets?
Put the weights in one row, say B2:E2, and each option's scores in the rows under it. The weighted total for the option in row 3 is =SUMPRODUCT($B$2:$E$2,B3:E3); fill it down and sort by that column. SUMPRODUCT works the same way in both. Or paste the totals from this calculator: Copy as CSV gives the weighted points, totals, percentages and ranks ready for a spreadsheet.
Is a decision matrix the same as a Pugh matrix?
They are close relatives. A Pugh matrix compares every option with one baseline option, marking each criterion better, the same or worse (+1, 0 or -1) and adding the marks, often with weights. A weighted decision matrix scores every option on its own scale instead. You can run a Pugh matrix here with the 0 to 10 scale by scoring the baseline 5 on everything and the others 4, 5 or 6.
Should cost score high or low?
High scores must always mean better, because the totals add up. So the cheapest option gets the highest score on cost, the shortest travel time the highest score on travel time, and so on. Write down what each score means before you start (for cost: 5 under $20,000, 4 under $25,000), so everyone scores the same way.