From Insight-less to Insightful: Four Strategies for Designing Impactful Data Visualizations
By rethinking your approach, you can increase understanding and innovation
As Alberto Cairo, a well-known information designer, professor, and the Knight Chair in Visual Journalism at the University of Miami, has long argued, data visualization is an efficient and important tool for helping humans learn from data:
“Most people grasp the truth of an assessment only when they unequivocally envision the evidence for it, something that our kludgy brains alone often can’t do well. That’s why visualization works.” - The Truthful Art
Data visualization is about visually encoding data so we can more easily recognize patterns. However, I came to realize that too often data visualization students and designers underestimate how powerful data visualization can be and limit the potential of their visuals, keeping them from finding unexpected insights. As I tried to get myself out of this limitation, I identified four fundamental strategies that can help data analysts and visualization designers transition to a higher level of design:
- Be the blasting cap of innovation
- Keep your expectations out
- Design for insight, not beauty
- Plan before you build
1. Be the blasting cap of innovation
Data analysis and data science are great tools to help businesses to make decisions based on facts instead of intuition or guesses. Data visualization plays a vital role in this process, usually by helping translate the results found in the data to decision-makers (or any other audience that you are trying to reach). But data visualization can do more than that! We can use data visualizations to not only answer questions but to generate more questions and inspire innovation. Data visualization can be the blasting cap of innovation.
But what is a “blasting cap”? Let’s think about dynamite for a moment.
Dynamite needs to be triggered by near-instantaneous heating to explode. This heating is called the “initial activation energy,” and it creates a small explosion that in turn detonates the dynamite. The activation is produced by a device called a “blasting cap.”
We can think of data visualization as the blasting cap producing activation energy, and the explosion of the dynamite as the resulting ideas and innovation. Data visualization can do more than answer pre-defined questions. By using data visualization to provide a new perspective on a topic, we can encourage people to think more creatively. New ideas come up, and innovation explodes.
Check out my article “Data Visualization, Let Them Wonder” for an in-depth discussion of using visualizations to widen one’s knowledge:
2. Keep your expectations out
One of the most common ways we keep visualizations from reaching their potential is by designing them based on the results we expect to see. I overheard each of the following in the past years while working with data visualization:
“Let’s use variables X, Y, and Z in our model so that we can show that our service is efficient”
“Please, remove those years from the axis for the report to support my project”
“Use this data set because it’s better to prove X and Y points”
What’s wrong with those sentences? When we build our visualizations around what we expect to see, we introduce bias into our analysis, and the whole concept of using data visualization to provide new perspectives and generate new ideas is broken. The insights your visualization provides will be limited or, worse, misleading. Make sure you are objective when designing your visuals, and focus on learning from your data.
3. Design for insight, not beauty
How many times have you seen a visualization that is truly beautiful but hard to understand?
Some years ago, Cogent Legal posted an article about showing complex data for litigation using Adobe Illustrator and Excel. One of the charts they presented looks pretty nice, with different colors, color intensities, categorical numbers, and a map. However, it is too complex — it’s difficult to know where to look first, what the different colors mean, and, especially, whether the map is just an illustration or if it is associated with the graph. Take a look:
Cogent Legal later revised the graph. They used different breaks along the x-axis, replaced the data points with a scatter plot and a line plot in one graph, added other data points in a different color, and added bars of increasing opacity at the peaks. The revised graph was even more complex than the first.
The chart still looks nice, but it isn’t easy to pull insights from. This could be helped by using multiple graphs instead of one. If you have a lot of information to share, try splitting it into multiple plots so that important aspects of the graph won’t overlap. To design for insight, keep things as simple as you can.
4. Plan before you build
Before you start creating your visualizations, take the time to understand the data and the focus of your work. Ask yourself questions that will give you a clear direction. For example:
Who is the audience?
What do you want to accomplish with your visuals?
What is the most important information for your audience?
What aspects of the data they need to understand better?
Then start exploring your data, and find what needs to be translated into visuals. This will help you tell your audience the best, clearest story you can.
I personally never start my visualizations on the computer, but instead with paper and a pen. I write down my ideas, goals for the project, and start sketching types of visualizations that could help my audience better understand the data.
A few months ago, when I was working on a dashboard about COVID-19, an idea came to me in a meeting. I immediately grabbed a piece of paper and drew what I had in mind.
That paper helped me, later on, to improve my ideas and decide on the look of my final dashboard. Stepping back to focus on planning before starting to build your visualization will help you to be more efficient, more creative, and take a broader, clearer view of your data.
As you plan your next data viz project, remember these four strategies to help you move from insight-less visualizations to insightful ones. They might be just what you need to turn your visuals into blasting-cap detonators and trigger an explosion of innovation for your audience!
- Be the blasting cap of innovation — use your visuals to change the perspective and expand the knowledge of your audience
- Keep your expectations out, and focus on learning from your data
- Design for insight by keeping things simple
- Plan before you build
