Do we need to follow brand colors for data visualization?

Question

I am working on a design language project for our product. I'm currently putting together a color palette for data visualization part of the app. Based on my research on colors for visualization and careful consideration of color theory, and accessibility, I came up with a set of colors (muted color palette) that would blend well with the UI. I worked it off the brand colors. Because our primary brand colors are Orange, Grey, and Green and ours is a typical network/device management app.

So, the question is: do we need to use brand colors for data viz?

Of course, the product UI conforms to the brand colors. However, I feel the chart colors not necessarily be restricted to the brand palette for a couple of reasons:

1. The marketing team which originally came up with the brand colors would not have thought about colors that would work well for Data viz. And I think they are more of a brand guideline that's suitable for the promotional/marketing material and for the corporate website. They don't seem to fit the bill for chart colors. We do have secondary and tertiary colors. So some of the colors I've come up with are derived from the secondary and tertiary colors and I've reduced the saturation level so that it's easy on the eyes.

2. It is an NMS app, orange and grey (primary brand colors) has some meanings in our app context (e.g., error, severity, others/unknown etc.,)

Should the chart colors represent the brand? Are there any guidelines on this? If someone could point me to some best practices, it would be helpful.

Answer 1

The colors of the chart should not represent the brand. Now I cannot find any study / research on this but it derives from logic that in your case you might have a 2 - 3 color chart and you are thinking of incorporating the brand colors to your charts, but once you get into complex charts with 4 - 10 colors you will not be able to use the brand colors.

What we did was define chart colors in the language system by 3 types of charts: - Sequential ( heat maps , tree maps ) - Categorical ( bar charts,line charts ) - Polarized ( Extremes visualised in heat maps )

and we ended up with around 20 colors and shades or tints.

Answer 2

The UI should work with the logo or branding and reflect the brand. The content should be separate from the UI visually. You can get that separation through other ways like saturation but it doesn't sound like that's the case.

Data visualization is like part of the UI but it gets its own colors.

It's a really broad subject so here's some backup and explanation of this in Material Design guidelines.

Material Design Color System. This describes their primary UI color system and there's a section on Additional Colors for Data Visualization

Material Studies applies color rules to an example data visualization app Rally. You'll see how they use extra colors for data visualization componenents.

Answer 3

There are two options: For continuous data, you can use a single hue color palette, and for categorical data, you use the multi-hued color palette. Make a 2-3 color chart palette if the data you present is not much complex. Or else you can build up to 6-8 color chart palette if the data has a verity of the filters like - browser, state, country and days with the comparison filters.

You can check the mixmpanel analytics tool data visualization color palette: https://design.mixpanel.com/#data-visualization-palette

Answer 4

As you are transformin specific information in it's own context and put that into your content strategy, info visualization should always represent the colors of the content.

It would look strange if the New York Times would color the election results in their brand colors. Even more weird if you are from a country where you have 5+ political parties. This would work for all other things, too.

As example, if you are talking about renewable energy, you should use colors like dark blue/ or yellow for solar, light blue for water, grey for wind etc. This way the user can subconsciously connect the abstract visualisation to the actually content.

TL;DR: No, usually you do not use brand colors for information visualization.