When your data is in percentages, think about pie charts. Pie charts make more sense when your raw data is in percentages, unless you have too many observations that clog the pie, in which case you’ll go back to a bar or column chart. Pie charts that display the proportion of the pie consumed as opposed to the quantity left are the most useful. Pie charts are vibrant and appear to display a lot of information, but they are typically employed in a way that does not improve understanding over a stacked bar chart, for example. It is only useful for displaying the relative contributions to a sum of several components when there aren’t many components and the relative amounts aren’t particularly high.

Pie charts have a somewhat limited range of applications, which is best summarized by its definition. You need some sort of full quantity that is separated into several different portions in order to utilize a pie chart. Instead of comparing groups to one another, the main goal of a pie chart should be to compare each group’s contribution to the total. A different plot type should be used in place of the pie chart if the aforementioned requirements are not met. Here is a pie chart I made for the most popular sodas in the US. And here is where you can get that data.