While a chart visually represents a numerical data set, labeling is critical to providing perspective and promoting legibility. A chart without labels is pretty much useless. So while it may seem like a small issue, it’s important to get right because bad labeling can easily lead to confusion, reflecting poorly on the presenter.
The two primary ways in which data is labeled in most charts:
- Axes: These are reference “lines” that define the coordinate system of the plot area.
- Direct labeling: Placing data labels on the items, such as bars or lines.
While almost all charts include one or both of these approaches, they operate by different rules (or should!), specifically when determining number format.
As your friendly neighborhood “chartist”, this is what to keep in mind when applying labels to your charts.
This is not my first missive on axis labels as they are all too often poorly executed, but I’ll keep my point brief and focused on issues of magnitude. When creating axis labels, the significant digits after the decimal point should be limited to the absolute minimum, which is exactly enough to accurately process each label. My personal view on this is that the significant digits following the decimal point should be the same as the label step size. So a step size of 0.1 limits the format to one significant digit.
Trailing zeros after the decimal point should never be added to the axis number formats, even if the underlying data is shown to a finer degree of accuracy. This trailing zero rule even takes precedence over general quoting conventions, such as the tendency to label dollar and cents with two decimal places. For example, a stock price time series chart may have underlying data significant to two decimal places (e.g $21.35) but if the axis ranges from $0 to $100 at increments of $10, no decimal places should be shown (e.g. $10, $20…$90, $100). Similarly, if the chart axis ranges from $0 to $1 at increments of 10 cents, the axis labels should show one decimal point (e.g. $0.0, $0.1…$0.9, $1.0).
Comparatively, data label formats are useful when you need to communicate greater accuracy as they should reflect the number of significant digits, market conventions, or subjective oversight and provide greater granularity than axis labels. So all those decimal places you didn’t show on the axis label, you can show here.
For example, using the chart above, the axis labels have no decimal places due to a 10% step size, but the bar data labels are shown within one decimal point. In this case, one decimal point balances nicely between accuracy and legibility, but that may not always be the case. While the data itself is a ratio and could be displayed with any number of decimal places, rounding off at one provides enough precision to compare and distinguish the individual data points, while keeping the chart readable. It’s subjective for sure and as a general rule, I wouldn’t consider the use of data labels as permission to go crazy with additional digits, as always with data visualizations, you need to strike a balance between clarity of idea, accuracy, and visual interest.
Questions about data visualizations, creating content, or curious to know how much more I can say about axis labels (answer: more than you would think), email me at email@example.com.
Explore Pellucid’s content, created specifically for pitchbooks and financial analysis. Request a demo.