A Pie/Doughnut chart expresses a part-to-whole relationship in a data set. It will always add up to 100%. So, if, for instance, you want to show "How does 100% divide up into multiple shares?" you should always use a Pie/Doughnut Chart with % as a data label.

1. Pie Chart

• In Chart B, we can quickly identify that the 25-34 group has the highest 'Users' share, but we can only guess the % share for any of the age groups.

• Unlike a Scorecard, a Pie chart doesn't have a Metric name. In Chart B, we can't see which metric is used.

Recommended: Yes [A]

• Yes, this is one of the most widely used charts across the industry.

• If you want to show "how does 100% divide up into a few shares?", it's best practice to use only Pie/Doughnut Charts.

2. Doughnut Chart

• In Chart B, we can quickly identify that the 25-34 group has the highest 'Users' share, but we can only guess the % share for any of the age groups.

• Unlike a Scorecard, a Doughnut chart doesn't have a Metric name. In Chart B, we can't see which metric is used.

Recommended: Yes [A]

• Yes, this is one of the most widely used charts across the industry.

• If you want to show "how does 100% divide up into a few shares?", it's best practice to use only Pie/Doughnut Charts.

• My personal favorite, the advantage of using a doughnut chart has a space in the middle for a Scorecard. In this example, "The total users making up 100%". This way, audiences can compare against an individual element.

3. Pie/Doughnut Chart with Value as a Data Label

• In both Charts, we can see that the 25-34 group has the highest users, but we can only guess the % share for any age group. Your report audience will always expect more information from a Pie/Doughnut chart.

Recommended: No

• If you want to see a part-to-whole relationship in a data set, chart data labels should always be % and avoid having number labels at any cost.

4. Column/Bar Chart as a Pie Chart Alternative?