Questions
Weight the criteria before you meet the options
Criteria weights say, in numbers, what you care about and how much. Most people set them by feel, and that is where many decisions quietly go wrong. Here are three reliable ways to set weights, the strengths of each, and a check that catches weights that do not match what you actually value.
Divide 100 points
The simplest method: give yourself 100 points and divide them among the criteria. The fixed total forces trade offs, since every point given to price is a point taken from something else.
Its weakness is that people spread points too evenly. With six criteria, many will give each one between 10 and 25, which flattens real differences. If your top criterion would decide the choice on its own, its weight should show it.
Rank first, then convert
When people find numbers hard, ask them to rank the criteria instead, then convert the ranks into weights with a fixed formula. One common formula, called rank order centroid, gives three ranked criteria weights of about 61, 28 and 11.
Ranking is quick and easy to agree on. Its limit is that it assumes the gaps between ranks follow a set pattern. If your first two criteria are almost equal, adjust the result by hand.
Swing weighting
Weights should depend on how much the options actually differ, not on how important a criterion sounds. Safety sounds more important than parking, but if every option is equally safe, safety should carry little weight in this particular choice.
Swing weighting captures that. For each criterion, picture moving from the worst option on the table to the best. Ask which of those improvements you would most want, give it 100, and score the others relative to it. Then scale the results so they total 100.
Weights in a Decision Room
Argumont asks for criteria with weights that total 100 before any option is scored, so the weights cannot drift toward a favorite once the options are known. The Challenger can press on a weight that seems to carry the leader on its own, and the Judge names in the Verdict the conditions, sometimes a single weight, that would change the result. The Human Decision stays with the person who owns the choice.
Swing weights for an office move
A firm compares three office sites. Rent ranges from $38 to $52 a square foot, staff commutes differ by up to 25 minutes, parking ranges from scarce to ample, and all three buildings meet the same accessibility standards. Asked which swing matters most, the partners choose the commute and give it 100, then rent at 70, parking at 30, and accessibility at 0 because no option differs. Scaled, the weights are commute 50, rent 35, parking 15. Accessibility stays a requirement every site must meet, rather than a weight.
Questions
- Should different people weight the criteria separately?
- It helps. Have each person set weights alone, then compare. Large differences show where the group disagrees about values, which is better discovered before scoring than after.
- How many criteria should I use?
- Usually four to eight. Fewer can miss something that matters. More tends to produce many small weights that add little and make the result harder to explain.
- Can I change the weights later?
- You can, with a written reason, ideally before seeing how the change affects the ranking. A weight changed after the scores are known deserves extra scrutiny.
Bring Argumont a Decision worth challenging.
Start with a paid 60 day pilot on your own real Decisions.