I have been writing about pie charts since 2009 when I wrote how pie charts are bad and I haven’t stopped writing about pie charts since then. Rather than a critique of some random pie chart I have seen, or been sent, or been linked to by the Chicago Tribune – I wanted to talk about how pie charts don’t belong in the board room.
The data in question is not necessarily in the board room, it is basically the data people are using to make decisions. There are a lots of different things you might measure at a point in time, but no matter what that data is – it needs two important properties if you want it to lead to a good decision.
The first property of a solid decision-making metric is trend. It literally doesn’t matter what the distribution of mobile operating systems was at a single point in time. What matters is the trendlines. It doesn’t matter that Windows Mobile had a 15% US market share in 2009. What matters is that this was down from 27% on the previous year, and from 49% in 2007. The trajectory of the market share tells you more than the actual number.
The second property of a solid decision-making metric is relationships. When you overlay trends for multiple things, you can find relationships. That can mean showing the relationship between the Windows Mobile market share, and competitor market share. But it could mean the relationship between Windows Mobile market share and your business revenue.
As you can probably tell if you have ever looked at a pie chart; they aren’t much good at showing you trends or relationships. So if you are using pie charts in the board room – you could be making terrible decisions. While line charts and column charts may lack the big-blocks-of-colour effect, they will help you be smarter for your business.