When an individual creates a graph, they make a ton of different choices—many intentional and some perhaps not—that influence the overall design. This probably sounds obvious. But I think it is an important point. No “expert” is “right” when it comes to the many specific design choices that go into visualizing data.
Data visualization sits at this fun intersection between art and science. Science in the guidelines and best practices to follow but also artistic in the many choices one gets to make (I should note that not everyone agrees—to read a lively debate on this topic, check out the 2016 exchange between Stephen Few and me on his blog and mine).
The end result is that two different people can approach the same data visualization challenge in completely different ways. That’s ok. That’s actually great. It means there’s room to put some of you into your data visualizations. This may be through your personal style or a way to let the branding of your client, team, or organization shine through. The personalization can take on so many different forms—preferences for certain fonts or color combinations, decisions on whether to keep or remove that axis line and tick mark, or even proclivities for (or against!) certain graph types.
As I build my team at storytelling with data and add more people to the mix, I’ve been reflecting on this idea a lot lately. I’m trying to determine when data visualization should be done in “Cole’s style,” and when someone else’s personalization will have its place. I don’t have an answer yet. But as part of my reflection, I thought it would be useful to take a step back and define and describe the basic principles that shape my work. These concepts will be familiar if you've read my book or attended a workshop, but they are framed slightly differently here, in terms of seven guiding principles for making an effective graph:
- Be clear on your intent. Are you visualizing data for exploratory purposes (to understand it better) or explanatory purposes (to make a specific point clear to someone else)? How you want your audience to use your visual has important implications on how you design it. In cases where you are explaining—nearly always the case in my work—have a point of view and make it clear.
- The right graph creates an “aha” moment. You know you’ve created an effective graph when it allows you to see an insight that was not apparent before. The graph should make the right way to look at the data feel intuitive for your audience; it shouldn’t feel like work to look at it.
- Don’t overcomplicate. This isn’t about over-simplifying or “dumbing things down,” but rather not making things more complicated than they need to be. It’s relatively easy to take something simple and make it complicated. It’s much harder to take something complex and present it in a way that is accessible to your audience. Aim for straightforward.
- Get rid of the non-essential. I’ll refer to one of my favorite quotes by Antoine de Saint-Exupéry, “You know you’ve achieved perfection, not when you have nothing more to add, but when you have nothing to take away.” Identify and eliminate clutter and distractions.
- Make it clear where to look. Don’t make your audience process everything to figure out where they should pay attention, rather direct their attention strategically to the most important part(s) of what you are showing. Create visual hierarchy by emphasizing some elements and pushing others to the background.
- Words make a graph accessible. There are some words that must be present: every graph needs a title and every axis needs a title. Exceptions will be rare. Title effectively and consistently so your audience doesn’t question what they are looking at. Include a footnote with your data source, methodology, and to spell out any acronyms.
- Audience trumps all else. Ultimately, everything we are doing when we communicate with data is for our audience. Keep that audience in mind throughout the design process and above all else, make what you are creating meet their needs. In some cases, it may mean softening on some of the fore-mentioned principles.
When we think back to the personal style that I mentioned earlier, each individual might approach nuances within the above differently—make different design choices than I might personally. But I believe if you follow these principles, you will be on the right track for creating effective data visualizations.
I’m interested in your feedback. When and how can personal style influence data visualization design? Are there guiding principles I should consider adding or modifying? Or others you see in my work that I’ve failed to include here? What are your guiding principles when visualizing data? Leave a comment with your thoughts!