Tell me if this story sounds familiar. You and your business partner are having a disagreement. He wants to heavily discount the price of your product to customers in certain parts of the world. In order to enter some promising emerging markets, particularly in China and India. You don't like the idea. You're concerned that your existing regular customers deserve best pricing as well, and they are likely to press you for the same discounts. The discussion gets heated. Then later you check your cash flow models to test the scenario your partner had suggested. And you discover that the actual discounts proposed by your partner have very little impact on your profits. But on the other hand, the shipping and transaction costs are actually too high to preserve your profit margins. In other words, you two have been fighting about the wrong things. The cure for that problem is sensitivity analysis. Sensitivity analysis helps us identify which variables in our models have the most impact on our desired outcomes. In this lecture, we will learn how to use spreadsheets to conduct what-if analysis with our models. We will examine a couple of ways to determine which variables are the most sensitive and therefore require the most attention. So now, let's turn to the spreadsheet. Sensitivity analysis is a disciplined approach tot he what if analysis we were conducted earlier. Sensitivity analysis asks the question what amount of the change in the desire to upload a variable is driven by other variables in the model. And in particular by our decision variables. For example, if we're concerned about improving our margins in the innovative speaker's cash flow projection. We could play with changing the price, or doing a better job reducing returns or increasing advertising to increase growth. The question is which one of those is the best thing to do to improve our margin. To test that we're going to increase each one of these three values for return rate, annual sales growth rate, and unit price by 10%. So in the case of a return rate what we'd like to do is actually reduce from 10 to 9%. And when I do that I notice that the margin which is currently at 15.5% changes to a little bit better than but not very much, 15.6. >> To check sensitivity, return to the base state, the value of the return rate, and check a 10% change in the next variable. In this case, annual sales growth rates. It's currently at 20%. If I take it instead to 22%, I can see how that affects the margin, currently at 15.61, and it does increase it by about a point to 16.51, making it more sensitive to the annual sales growth rate change than the return rate. But meanwhile, if I return again the sales rate, its growth rate to its base state and change the $89 to $99, [INAUDIBLE] 89 times 1.1. You'd notice if I had a much more dramatic impact on my margin. The margin has now risen to 20.57%, indicating that our margin is much more sensitive to changes in unit price than to either changes in the growth or the return rate. So what we're saying is given limited time and resources, the model is telling us to focus on increasing unit price. And testing market acceptance of those higher prices. This type of sensitivity analysis is sometimes referred to as, one at a time analysis or OAT. There are tools available both free and fee based for further use of this technique.