How sensitive is your answer to one uncertain input?
An estimate can look precise while depending heavily on a guess. If an event needs one volunteer per ten guests, an attendance estimate of 50 implies five volunteers, while 100 implies ten. Sensitivity analysis changes an uncertain input to see how much the resulting decision changes.
Go deeper
Test the estimate around its weakest input
A planning sheet calculates staffing from expected attendance. At one volunteer per ten guests, fifty guests require five volunteers and one hundred require ten, under the assumed rule.
If attendance is poorly known, a single staffing number hides the uncertainty. Sensitivity analysis varies that input while holding the rest of the model constant.
The exercise reveals whether a small change produces a small adjustment or a major shift. It can also show which assumptions deserve further checking.
This is not a claim that every tested value is equally likely. A range is a way to inspect the model's behaviour. The staffing rule itself may also need review if different activities require different levels of support.