Every comparison rests on assumptions, such as how much cost will matter to you or what prices will look like next year. Most are estimates, and a few are honest guesses. Estimates and guesses are fine, as long as you know which parts of the result depend on them.
Take an answer that tipped the result, such as your answer about cost. If you change that answer a little and the same option stays on top, the result doesn't depend on that answer. If a different option jumps ahead, the decision hinges on that answer, so that answer deserves careful thought.
Facts work the same way. If one option's lead rests on a figure that might be out of date, checking that figure before you commit is worth the time. A Decidient result shows which of your answers the leading option is most sensitive to, so you know where to start.
A result that holds up under reasonable changes is less dependent on those assumptions. A result that flips easily means the choice deserves closer attention.
The science. Testing how a result changes when the inputs change is standard in decision analysis. Lawrence Phillips argued that a model is ready to act on when more testing stops turning up new ideas (Phillips, 1984).
This page was written by AI from the sources cited here and reviewed by a person. Last updated 1 October 2026.