A sensitivity analysis is a forecast of which input factors to a financial model will impact the output of the model the most. The output can be any Return on Investment indicator, such as Net Present Value (NPV), Breakeven, or actual Return on Investment (ROI). In many cases, you may need to capture more than one output as you are testing input factors. All Models are Wrong, Some Are Useful. - George E Box, one of the great statistical minds of the 20th Century. One of the major criticisms of financial models is that they "are always wrong." And it's true. Just about everything in a model is a Wild Ass Guess (WAG). But that is because most business planners don't take their models to the point of understanding how bad is it if it's wrong. What if your costs are 20% higher than you expect. Does the business fall apart? Can you accept that risk? Statistical sensitivity of financial models can tell you how bad it is if you're wrong. Statistical sensitivity tells you which of your WAGs is the most important to get right. With statistical sensitivity, your WAGs become SWAGs (Statistical Wild Ass Guesses). SWAGs make you look smart, you want to make SWAGs, not WAGs for your business. Statistical sensitivity analysis turns Wild Ass Guesses (WAGs) into Statistical Wild Ass Guesses (SWAGs). You want to be good a SWAGing in business. - Nate Kaemingk