One of my favorite outputs from our marketing analytics models never gets its due. It shows which factors are moving the business results more than others. For instance, if I were running an online retailer I’d certainly be interested in hearing whether my sales are more sensitive to consumer confidence or gas prices. Enter the Tornado Chart:
In the above example, we see that if this retailer were to grow their store count by 1%, they would see a 0.82% growth in sales. Why not 1%, you might ask? In this case, it is likely that they have already taken all the best locations, and new locations are not likely as high quality as their store base. Another interpretation that happens to large retailers is that they start to cannibalize other nearby stores in the area.
The red bars also jump out from the page: they represent factors that have a negative relationship with sales. So when unemployment goes up, sales go down.
The best use of the chart is for a high-level viewpoint of what is really moving the needle. Our clients who have multiple products or multiple channels find it extremely useful for comparison and broad business planning. They also pay attention to competitive impacts- if they find a product set that is especially sensitive to competitor marketing they use it as a proxy for product or category assortment weakness.