Which chart do I use?
When trying to figure out which chart to use, ask yourself who your audience is and what you want the data to show. Sometimes a bar chart really is the best chart to tell the story.
Charts can be categorized by four different objectives: comparison, correlation, distribution and composition. This is not a one-size-fits-all solution to picking the right chart. Designing effective, story-telling products is more than just choosing the right chart for the data set. This Chart Selection guidance should help you get started when trying to decide the best chart for your data.
Comparison
Do I need to see the difference or similarity of metrics within the data?
Charts that show comparison between data sets show patterns and trends quickly. Comparison shows similarities or differences. This chart function helps to highlight patterns and outliers.
Other charts: Pie/Donut, Heatmap, Stacked bars, Bubble, Dot plots, Waterfall, Butterfly, Lollipop, Waffle
Correlation
How do these metrics relate to or affect each other?
Charts that show the connection between two or more variables in the data or how one variable might affect other variables.
Other charts: Sankey, Heatmap, Dumbbell, Venn
Distribution
How do these metrics tend to behave, either over time or in relation to themselves?
Distribution can show the frequency and how a specific metric spreads out or clusters. A distribution of a metric plots a collection of information to find out how it spreads and to understand if there’s any interaction between the variables.
Other charts: Line, Treemap, Heatmap, Bubble, Scatterplot, Boxplot, Stacked Area, Area, Dot Plot
Composition
How do the individual metrics make up or compare to the whole metric?
Composition charts present each individual metric as compared to the whole
Other charts: Progress bars, Bullet charts, Bullet, Waffle, Marimekko, Gauge
