As discussed here, you don’t have to label every single slice in a pie chart. Depending on the message, you can choose to label only the more relevant ones. Let’s see how this can rescue a really bad chart.
This is a well-known chart from the Wikipedia that displays the total population for each of the states in the US. There are a few issues, like the “Other” at the top, but the main reason why this pie is known is because it contains perhaps 10x the number of recommended slices for a pie chart. But is that the real reason?
I don’t think a pie is a good chart to display several dozen values, but this example is poor because it tries to force the chart into the wrong mode. With so many slices, you have to focus in a few key data points that will support your message.
Let’s try to improve the pie chart using a similar data set. The two pies below display the same number of slices, but interesting stories are easier to find. The message in the first one is clear: four states account for a third of the population in the US. That’s all it wants to say.
The second pie groups states by region and there are interesting stories to tell, like the fact that the South accounts for more than a third of the population, and that, if you add California, that’s exactly half of the population.
For us right now it doesn’t matter if these examples make sense or are useful. What matters is that, if you think a bit more about the data and the story you want to tell with it, even a seemingly hopeless chart can be redeemed (to some extend).