Designing Excellent Charts: Show and Don’t Tell
This week we published Versta Research’s quarterly newsletter with a feature article entitled “How to Design an Excellent Chart.” It addresses one critical piece of turning research data into compelling stories by focusing on the process of data visualization. For us, that process involves five steps:
1. Determining whether a chart is really needed. A lot of data is more easily communicated via tables, and sometimes citing just one or two numbers in text is more effective.
2. Identifying the right contrast to show. A good chart will lead viewers to make certain data comparisons versus others, so it is critical to determine the story the chart needs to tell.
3. Considering multiple types of charts. Bar charts, pie charts, and line graphs all function in ways that tell different stories. We always experiment with a few variations to see how each works.
4. Telling the story. We test the chart on colleagues to verify that it tells the right story without help from a narrative, and then we add that narrative to tell it as well.
5. Adjusting based on best practices. There are do’s and don’ts about colors, shading, labeling, graphic design, and so on. We incorporate as many as we can so that they become second nature.
We take inspiration from Ernest Hemingway’s advice to “show the readers everything, tell them nothing.” A great chart helps open up the story and show what is implicit in the data. We also share some news stories and media coverage of Versta Research’s recent work, including a national survey for Fidelity Investments of workers with pension plans, and a survey of knowledge workers at Fortune 1,000 companies to identify key moments that can make or break employee engagement.
We hope you find useful and thought-provoking ideas in the newsletter. Let us know if there is anything we can do for you.