As we mentioned in our first Digital Trends Report last month, the rise of big data in marketing research has made visualizations crucial to how brands perceive and present themselves. Although data visualization — including the ubiquitous and questionable “infographic” — should involve training and careful consideration, unfortunately, expertise and good judgment do not always prevail.
I recently came across a Tumblr focusing on exactly this: bad data visualizations. Although the examples are hilarious (the URL is viz.wtf, after all), the site isn’t just for laughs: These examples are also educational, providing a great reminder of what not to do.
Here are some of my top #dataviz pet peeves (all images from viz.wtf) and some tips for how to avoid them.
Data visualizations often suffer from unneeded flourishes that the designer (or lack thereof) thinks makes the presentation more attractive. 3D charts are a repeat offender, as is the case with the chart to the left.
Lesson: In data visualizations, every design aspect should have an intended purpose. Nothing should be for appearances alone.
Good visualizations must strike a balance between information conveyed and simplicity of design. Although both aspects are equally important, they are rarely balanced. Most often, the creator opts in favor of more information, making the visualization busy and difficult to understand. The example at the right is seriously guilty of information overload.
Lesson: Minimalism means using fewer elements (and often, doing more with those elements). In visualizations, graphics and charts should be simple and to the point, while working together to tell a comprehensive story in a few brief glances. Yes, this is very difficult to do well.
More often than you might believe, visualizations contain charts and graphics that make absolutely no sense, violating principles of mathematics or statistics either out of ignorance or to achieve a certain look. My favorite example from viz.wtf is the man who is “243% Baby Boomer.”
Lesson: Never manipulate data for aesthetic reasons: Get to know your data before deciding what you want your visualization to look like. Also, have a basic understanding of how percentages work.
The most dangerous flaw behind data visualization is not having a central question to answer. The question (and answer) are the beginning and end of your story, and without them your story is confusing or boring (maybe both). Every project, report, presentation, or yes, visualization should have a business question it tries to answer.
Lesson: Tell a strong story. Every graph, table and chart should move the story forward the same way a character in a movie does. Use your characters wisely and arrange them in a logical narrative order.
Pro tip: Geography and time lapse data, when available, tell the best story (because stories take place over space and time, you see?)
Examining data visualizations that didn’t work well is certainly a fun way to learn, but even more value comes from looking at good examples of visualizations. In addition to our friends at viz.wtf, I recommend checking out the following #dataviz influencers and producers:
And of course, get in touch with us if you need help making sense of it all. Delucchi Plus recently rolled out our Insights product group to do exactly this: Help you and your company make the most of your marketing research investment by providing creative, actionable insights.