Remember when those magical words “once upon a time” promised an exciting story? We strive for the same excitement, or level of persuasion, when storytelling with data but we often miss the mark. Ed Stalling, our Chief Storyteller at Maritz Research, often talks about his dream that at the end of a presentation, the audience would leap to their feet in dance similar to a chorus of “Lechaim*” from the Broadway musical Fiddler on the Roof.
The closest I’ve come dancing after a presentation was a couple weeks ago. I was in the audience of a Maritz customer experience summit that a good number of auto manufacturers attended. Cole Nussbaumer, founder of Storytelling with Data, shared her approach to bringing data to life by focusing on simplicity and ease of interpretation. Cole honed her data visualization skills over the past decade through various analytical roles and was most recently at Google where she developed and taught a course on communicating effectively with data.
I was impressed with her discussion on how to be a good storyteller. She drew upon some of the recognized authorities in the field of data visualization or infographics such as Stephen Few, Colin Ware, Dona Wong and Edward Tufte. She shared some great examples of how to bring data to life, what to avoid and why.
I don’t think I can do her presentation justice with just words but I’ll try to capture the essence of her message.
- Understand the context
- Choose the right type of display
- Eliminate clutter
- Draw attention where you want it
- Tell a visual story
- Practice makes perfect
Here’s a simple example I created based on concepts 2, 3 and 4. In this fictional situation with made up data, I’m showing how you might tell a story about whether the power of suggestion results in people using 3D charts. In other words, when you see these options below in Excel, are you more likely to add color and choose a 3D chart.
If you chose a 3D pie chart to show the data, you can see it distorts perspective and it is difficult to see the size of each piece of the pie. From this angle, it looks like “Excel uses them” is the largest section.
The right type of display is important. If you remove the 3D, it does help you see that “I’m not sure which chart to use” represents slightly more respondents. However, ask yourself if a pie chart really helps you understand the relationship between these responses.
A stacked bar chart helps not only illustrate that “I’m not sure which chart to use has slightly more responses than “Excel uses them in their icons” but a horizontal bar chart helps you see the relationship between the choices. (#2) And it eliminates some clutter since we don’t need a legend. (#3)
Keep in mind that my story revolves whether the power of suggestion results in 3D charts. Therefore, I might want to highlight the response “Excel uses them in their icons” to draw your attention to it rather than use color (#4).
Another consideration to draw attention to my story would be to reverse the order of the bars. In a final step to eliminate clutter, I’m going to remove the grid lines, y-axis tick marks, label each bar and add an axis (#3). This is a big improvement over the first pie chart. If I were including this chart in a presentation, the last step would be to include appropriate text to highlight my story.
Ed saved me from 3D charts years ago. As a result of Cole’s presentation, I’ll be better able to tell a story and use charts to my advantage.
What are your pet peeves or best practices about the use of charts to help tell a story?
You can access Cole’s site StorytellingWithData.comor you can follow her on Twitter at @storywithdata.
*Lechain or L’Chaim in Hebrew is a toast meaning “to life.” Ed and I are big musical theatre aficiaonados so I take any opportunity to add a reference to my favorite musicals.
