What is a Donut Chart?
A donut chart, in its simplest form, is a pie chart with its center cut out to look like a donut. At first glance, this may not seem to serve a much greater purpose than to create aesthetic variety. However, a donut chart helps avoid confusion around the area parameter that often trips people up in a pie chart.
When looking at a pie chart, it is easy to confuse the area of each slide with the entire pie and make deductions based on this visual cue. Instead, in a donut chart, the center is removed, which encourages the reader to focus on the length of the arc instead, and not compare it with the total area a circle would represent.
Another visual benefit of a donut chart is that the space inside the donut can be used to represent data, labels, and such to make reading the chart easy.
Each slice of a donut chart represents different chunks of data, often color-coded for easy understanding.
Why Use A Donut Chart?
In most cases, a donut chart can replace a pie chart since their use-cases are not vastly different. A donut chart typically shows the proportions of categorical data where the size of each piece of the donut communicates the proportion of each category.
The simplest example could be the extension of an elementary math problem. Assuming we have twenty apples, ten bananas, fifteen oranges, and seven pineapples, which fruit has the highest proportion in the basket? The donut with the longest arc represents the fruit that is present, as it happens, in abundance.
But to say that these are the only uses of a donut chart would be too simplistic. In general, a donut chart occupies far less space than its pie counterpart, which makes it ideal for placing in dashboards that already have a lot of information to process, as well as in business reports where data needs to be visualized simultaneously to draw the best insights.
For example, a donut chart of electronics sales for a company in 2020 can be placed inside a donut chart of electronics sales in 2021. Doing so allows one to place the data table right next to the chart and communicate more information than would be possible by placing two charts next to each other, and their corresponding data table underneath.
How to Turn Data into a Donut Chart
A donut chart works on two dimensions, and therefore, it needs two sets of information with a common relationship. The first dimension represents the attribute, and the second one represents its value. For example, in the electronic sales donut chart, the number of televisions sold is the dimension while a number (such as three hundred) is its value.
Both dimensions are important because they share a relationship. For someone reading a donut chart of electronics sales, it helps to know how many items were sold across which categories. Moreover, comparing with donut charts of other years also helps them see if item sales have significantly shifted.
One of the most common questions, or queries, raised about a donut chart is whether the values in the raw data need to add up to 100 percent. The simple answer is yes. A donut chart is best used to represent parts of a whole, but this is not to be confused with adding up to the number hundred. Any value, even one that runs into a few million, can be represented as a donut chart, as long as all of the values in total represent parts of a whole or 100 percent.
With most software products, a donut chart can be created by selecting the entire data table and choosing the donut chart option. Once done, values are usually arranged from the highest (the longest arc) to the lowest (the shortest arc) from the top clockwise. However, this setting is often flexible, and the chart maker can choose to rearrange these segments.
Benefits of Using a Donut Chart
A donut chart is one of the simplest representations of data and is also widely known. When presenting information to large and diverse groups of people, a donut chart is often the best option as long as the data being represented is a whole set, with multiple different parts within.
In sales reports, a donut chart can be used to study the number of opportunities that are open, lost, or won, and the resulting revenue can be represented as a donut chart. This allows decision-makers to see if the leads won contribute adequately to the bottom line, or whether the ones lost are too costly to lose.
To add to this functionality, a donut chart can be dynamic. Any change in data values can change the appearance of a donut chart, which can be useful when a forecasting exercise is underway.
Donut charts can be drawn in different sizes and colors, placed within one another, and labeled within the chart area to save space, making them more interactive and giving readers access to richer data.
Donut charts are considered to be one of the tidiest forms of data representation. Segments of a donut chart can be highlighted by making them thicker to represent segments that are of greater importance. Some software products also calculate segment percentages automatically when using a donut chart.
Simply put, those who are new to dashboards and reporting can use a donut chart with as much ease and pace as their more experienced counterparts, and yet manage to communicate the right insights.
Challenges When Using a Donut Chart
As with most other forms of data representation, a donut chart can become cluttered with too many segments. If there are too many segments to represent and each of them occupies only a small portion of total data, the segments can be hard to read.
A donut chart is not the best format to use when negative values need to be represented. For example, debt can be represented in a financial planning donut chart only as a portion of the whole and not in terms of its impact on total cash flow.
Also, while a donut chart is great for comparing data, analysis using only a donut chart is often hard because visual cues are the only means to read the chart. However, this issue can be solved by inserting the percentage labels next to each chart segment.
If changes over time need to be tracked, a donut chart is not the best chart option to use, because unlike a bubble chart’s tracker, it cannot retain the information it represents over a period of time. For example, a donut chart for annual expenditure continues to change as adjustments are made to its data source, without retaining any of the information from before the change was made. As a result, there is no way for the reader to see how expenditure has grown over time.
Alternatives to Donut Charts
A pie chart can replace a donut chart in nearly every application where the specific benefit of a donut chart—comparing areas as absolutes—is not a requirement.
Apart from a pie chart, the data in a donut chart can also be represented in other graph types:
Percentage Bar Chart
This serves nearly the same purpose as a donut chart. However, it can also be used when there is limited space, or when the total of all segments does not add up to 100 percent. For example, a percentage bar chart can be used to compare greenhouse gas emissions in 2010 with emissions in 2020, which would be higher.
A treemap offers a stunning visual tool that can be used to add more information about each of the segments. It often looks like a stack of blocks, and each block can then be used for adding a description with the data.
A Waffle Chart
This replaces a donut chart in cases where the data is larger and easier to comprehend when several squares are present as a visual aid. For example, it could be used to represent the number of people under age 20 per million people in any given year.
For those who wish to experiment with more interesting representations of data than using a plain table, a donut chart is a great starting point. However, it does come with its own set of limitations and must be used only when it provides visual benefit. For data that is segmented into very tiny bits or that often flows into negative values, other chart types such as a bar chart may be more ideal.
8 Must-have Capabilities to Take Your Data Visualization to the Next Level
Artificial intelligence, machine learning, and natural language processing are some of the analytics...
TIBCO Spotfire Data Functions
Data scientists are using Python and R to great effect—for data prep, feature engineering, machine...