Data Visualization-Elephant Data Visualization: Make Your Message Obvious

Many visualizations require extraordinary powers of deduction to understand them. Sometimes it feels like the author is intentionally trying to confuse you. However, it doesn’t take much to transform a graphic into something meaningful and easy to read. You don’t need to be a graphic designer or spend hours nitpicking your work. If you apply a few simple principles of data visualization anyone can understand what your data has to say.

The goal of any visual is to communicate an idea more quickly and effectively than can be done with words. There are two simple steps that can be applied to any visual to make its meaning obvious with minimal effort required from the viewer:  (1) Get rid of the junk, and (2) Make the big ideas pop.

Get rid of the junk

The first and easiest step is to remove all the clutter. It’s amazing how much you can take away from a chart without affecting its meaning. In fact, most common chart components are more distracting than anything.

Look at the following and see if anything is lost by removing most of its elements…

Data Visualization-Graph

Not only was nothing lost, but its clarity was significantly enhanced. And all I did was remove a bunch of stuff.

In general, if a mark is not displaying relevant data or giving necessary context then it needs to go. Examine every line or mark and ask if the ability to understand the chart is hurt by removing it. If not, delete it.

Make the big ideas pop

The power of data visualization is in its ability to communicate complex ideas quickly and effectively. A visual done well reveals insights that words and tables cannot. The difference lies in the ability to use visual cues to tell viewers exactly where to look. There are certain pre-attentive attributes that convey ideas more quickly than the conscious mind can think about them. Consider that for a moment.

What this means is that it’s possible to transfer information to a viewer instantaneously and with virtually no effort on their part.

This is truly remarkable, and it’s why leveraging the principles of dataviz can significantly boost your ability to communicate and persuade.

For a deeper discussion of preattentive attributes and how it applies to data visualization see Stephen Few’s Information Dashboard Design or Cole Nussbaumer Knaflic’s Storytelling with Data. Here are some of the most common preattentive attributes (inspired by the aforementioned Few and Knaflic)…

Data Visualization-hue

Consider this like a toolbox for making your message obvious to the viewer.

Putting it all together

So let’s put all these ideas into practice. I know you would never think about creating a chart as bad as the one below, but just for illustration let’s walk through how to turn it into something with a clear and easy to understand message.

Data Visualization-graph

To do list:

  • Chart junk:Remove unnecessary precision, colors, borders, and labels; deemphasize gridlines and axes
  • Visual cues:Reorder from smallest to largest; Give a splash of bold color to the largest region
  • Takeaway:Make a descriptive header

The final result is a simplified chart with a whole lot more meaning…


Now, I realize real world data and analyses are usually more complex than this basic example. But just because the process of discovering insights from data is challenging doesn’t mean the resulting graphics need to be. The concepts above can bring any idea to life no matter how complex and nuanced. In my experience, visuals fall flat because either the “insight” is itself flawed, or I wasn’t clear in my own mind on what I was trying to say in the first place.

A well-made visual can represent rich insights from large datasets and complicated analyses. Simply get rid of the junk and make the big ideas pop and your message will be obvious.


By: Dan Gastineau | aspirent |