While I find the waffle chart shares almost identical use cases to the donut chart, it excels at showing a more compelling way to visualize 0-100% KPI attainment than an isolated metric can. It also has a niche for showing parts-to-whole contribution. Other than these use cases, you’re mostly forcing things a bit. To me, the second scenario is the differentiator—a higher stressing of the individual points which make up the larger whole.
Good use cases
- Comparing high-level items or seeing how they contribute to a whole
- Placing increased emphasis on the part of the whole
- Using KPI charts as % of total
Bad use cases
- When many values contribute to a whole
- When KPI being measured could exceed 100%
- When exact percentages are vital, as showing fraction of a percent is more difficult to see beyond a rough approximation
Much like the donut chart, there are also two main variants: the same parts-to-a-whole and KPI options. I believe that waffle charts are much better suited for the 1-100% KPI variant than the stacked variant, unlike a donut chart which is a bit more flexible to do either.
In the example above, I’ve shown the stacked variant, where you can start to see some of the ways a waffle chart falls short. The primary issue is that comparing the sections labeled In transit (orange) to Delivered (Blue) is quite difficult unless you look at the actual number. In transit has a partial row on both the top and bottom, something we have to perform math to comprehend and compare effectively. Due to these issues, I prefer the KPI alternative, which is what I will walk through below.
This builds on a much earlier take on the same chart by our very own Jesse Gebhardt. The standard way to make waffle charts is often to round the values and have only whole squares shaded. In this example, I’m going show you how to make waffle charts accurate out to several decimal points.
I’m calling them “Precise Waffle Charts”, as I can’t find an actual name that denotes this distinction, but if you prefer the less precise, rounded variant, you can view and download a workbook with descriptions for each step here.
Keep in mind that even the precise waffle chart will never be ideal for extreme precision on its own because the human eye is unlikely to accurately guess how much of a square is shaded past a decimal or so. The label should be able to clarify the exact number to whatever decimal precision you prefer.