Pie charts are a popular visualisation choice used to represent a part-to-whole relationship. In other words, the ‘parts’ are proportional categories with a combined total of 100%. Its name is derived from how each section resembles a slice of pie. Pizza chart would be an equally appropriate name, but perhaps has less of a ring to it. Food references aside, pie charts are a default charting option in most visualisation tools such as Tableau and Power BI. However, those concerned with data visualisation best practice will know pie charts remain a contentious visual.
Pie charts often have labels of the category names, together with their relevant proportion and value. It is argued that this demonstrates why they are ineffective. They are difficult to understand on their own and rely on descriptive annotation. Moreover, one of the top criticisms for pie charts is the difficulty in interpreting angles and ranking the segments, especially if not sorted by highest to lowest. This further highlights the need for labelling .
Having said that, the jury is out on the effectiveness of pie charts and they remain commonplace when visualising data. As a rule, when displaying categorical data, always consider the number of categories and where possible filter/highlight the most important. When pie charts are used, they work best with minimal slices (a maximum of 3-4 would be advised) such as displaying binary Yes/No responses.
However, treating pie charts with caution, this blog offers five alternatives to consider when visualising proportions. Moreover, to avoid repetition it is wise to have a range of charts in mind when making decisions on the most appropriate visual.
Note: The visuals have been built using Microsoft Power BI and Tableau. The data represented shows the age breakdown (into 3 categories) of the UK population in 2018 .
1. Donut Chart
Donut charts are not too dissimilar from pie charts. The circle in its entirety represents 100% and each category is a coloured segment accounting for a proportion of the total. The signature hole in the middle resembles that of a donut, hence its name. It is the cut-out middle which arguably make donuts more favourable than pie charts. There is less attention on the specific angle, instead focussing on the length of the section along the circle edge. However, ease of interpretation can vary by individual and the ability to rank the segments remains a challenge. Additional labelling outside the visual would also still be advised.
2. Percentage Bar Chart
Percentage, or stacked bar charts, are a single bar with the width or height (depending on its orientation) representing 100%. Each category is a coloured section within the bar accounting for a proportion of the total. If you want to steer clear of pie charts, and the not too dissimilar donuts, bar charts are a great alternative. There are no potentially misleading angles to comprehend and the labels can sit inside the chart itself, providing an overall cleaner visual. There is also room for more categories than 3 or 4 as recommended for pie or donut charts.
3. Tree Map
A tree map is a group of rectangles inside a single rectangle. Combined, the rectangles represent 100%, with each coloured section accounting for a proportion of the total. Like the percentage bar, tree maps have labels inside the visual itself, and depending on the data, often do not require a legend (in this case the age categories sit nicely in each section). By default, each category will be ordered from highest to lowest, with the largest proportion in the top left and smallest in the bottom right.
4. Waffle Chart
A waffle chart is a 10×10 grid of squares with each square accounting for 1%, thus all squares total 100%. The signature squares are what give this visual its name, however, they look just as great with circles. Each category has the appropriate number of coloured squares/circles representing that categories proportion. Waffle charts are less common and may require some explanation. However, they can be a fun alternative and a great viz to have in your toolbox.
Note: this was built in Tableau Desktop but is not a default visual. Details of how to build waffle charts can be found here .
5. Bar Chart – Labels
The charts outlined above have kept the core principles of a pie chart: a visual representation of sections as a proportion of the total. However, this may not always be necessary. A simple yet equally effective technique is to use a bar chart and add the percentage as labels. You can order the bars from highest to lowest which clearly shows which categories account for the greatest/smallest volumes – a criticism often applied to pie and donut charts. However, in this case, the visual below has intentionally been ordered by age group instead of the proportion.
Pie charts are a controversial visual. If you like to use pie charts I recommend sticking to a minimal number of categories/slices. However, if you want to err on the side of caution, this blog has suggested five possible alternatives: donut, percentage bar, tree map, waffle, or simply adding labels to a bar chart.
Thank you for reading. Let us know your thoughts on pie charts and the different ways you have visualised proportions!
If you enjoyed this blog, check out [6 Alternative Visuals to Line Charts for Categorical Time Series Data].