I could write an entire post on color usage, but I’ll sum it up in three tips.
Color for Numerical Scales? Use with Caution.
This can get confusing. The way you interpret a shade depends on other colors around it.
Take a look at this map. If I ask you whether Idaho is doing better than Nevada, what would you say? Instinctively, you’d say Nevada’s doing the same, but the reality is that Nevada is doing better than Idaho. Nevada just looks paler because it’s next to California, which is doing really well.
We might be tempted to make a map with our data just because we have geographical fields and maps are so easy to make with Tableau. But, again, ask yourself the question: Does this help answer the question?
In the case of a map, they are really useful for looking at clusters, especially if the data contains geographical grouping. Perhaps we have fewer sales in northern states in the winter because it’s too cold, for example.
Maps tend not to do so well for simple comparison between states or regions for a number of reasons (we’ll talk about those next). But remember:
- Color perception, is, as mentioned earlier, a tricky thing. And having light and dark colors next to each other can lead to false conclusions.
- Choosing the right chart type for the question you’re asking is critical! The bar chart here answers the question of “which state is doing better?” far more effectively than the map.
Leverage Color Associations
Everyone knows that strawberries are red, pears are green, and oranges are, well, orange. Use that! And not just for everyday things. If you have a brand identity that your users strongly associate with, take advantage of that.
Color is an instant key that leverages long-term memory as soon as you see it.
We could confuse people:
Or we could make it so that people don’t even need the legend to interpret the figure:
Use Bright Colors to Highlight
Want to attract attention to a certain part of your data? Alarm colors draw the eye quickly to areas that need attention, and help get that message across.
Are our sales meeting targets? We can use visual best practices, minimize data ink, and create a bullet chart for easy comparison. We can color-code products that are meeting targets and those that aren’t:
The orange certainly captures my attention and tells me where I have problems, but that blue is pretty bright, too (and we’re using Tableau’s default color palette here, which was designed so that any one color doesn’t stand out too much). We want to draw attention to the orange. So let’s try a different approach:
Much better. The orange sticks out immediately; I know exactly where to take action.