Best Practice Guide
The ultimate step-by-step guide for analytic content creators.
Highlighting important data in charts
Where possible draw the users attention to the important data points in your chart through the use of various highlighting techniques. This can be achieved in a variety of ways depending on the analysis being undertaken from including trend and reference lines to highlighting outliers and anomalies.
Consider using a combination of the methods below if it helps you to communicate what your end users need to know.
Highlighting trendsUpdated 23 July 2021
If the intention of your visualization is to highlight changes over time then include trend lines or averages for a period so that users can quickly see the difference from one period to another if the changes the user is looking for is periodic.
<C73 Insert image – line chart, line chart with trend, line chart with average for 2 periods>
Increase the contrast between the series and the trend line, by lightening the series or reducing its opacity, so that the trend is emphasized. Alternately use an alternate color for the trend.
<C110 insert image, bar chart with average dark bars versus bar chart light bars, scatter plot dark dots, scatter plot light dots – see crime dashboard>
Highlighting outliers and anomalies
In a time series line chart if users are looking for unusual activity highlight outliers and anomalies. This draws the users attention to statistically relevant results not just guessing which periods may or may not have had a significant change.
<C74 insert image – line chart versus line chart with anomalies highlighted>
Ranking against peers
On a bar or column chart use an average line for ranking analysis. This line will help users to see which categories are doing better or worse than average. It’s a simple way to focus on higher or lower performers.
<C75 insert bar chart with average line>
Use color to highlight variation
Use color to highlight variation. Divergent color pallets are ideal when you want to draw the users attention to variance from target or a prior period. This makes the difference far more dramatic especially if good and bad values are encoded in the color choice.
<C76 insert image – bar variation no color, bar with color – red and blue>
Use reference lines to show targets
Reference lines are a good way to show the targets that need to be obtained. In this way users can easily see when or what is above or below a specific target. Reference lines should always be used for deviation analysis where the chart type is a line series.
<C77 insert image – bar chart no ref line, bar chart with ref line>
Use color to indicate attainment to targets
If targets have a stopped outcome – such as 100% is excellent, 80%-100% is good etc then use colors and conditional formatting to encode these rules and indicate how a particular value is tracking.
<C78 insert column chart – no color, with conditional color>
Highlight subject entities with color
If the analysis is showing how one entity compares to all others then you can use color to highlight that entity. This can be done in bar and column charts, line charts using monochrome for all other entities and data points in a scatter plot.
<C79 insert image – Bar highlighted, Line chart with monochrome, dot in map>
Use icons to convey meaning quickly
Use icons in composite visualizations to indicate change direction or good and bad. Icons can help a user to quickly identify the intent of the visualization without needing to interpret the data.
<C80 insert image – KPI no icon, with Icon>