Chart Design Principles
last updated March 9, 2017
Spot the difference between good and bad charts, based on this compilation of design principles from leading experts, with citations listed below to learn more.
Remember the most important principle: Find meaningful insights in your data, and create visualizations that help you tell these stories. All of the other details below are secondary.
Before you begin, ask yourself: Do I really need a chart to tell this data story? Or would a table or text alone do a better job?
Decide if the best way to communicate with your audience is with static charts (such as images printed on paper) or interactive charts (embedded in a website, with tooltip details and source links). Most of these principles apply to both types, but this book features tools and tutorials to create interactive charts.
Understand basic chart vocabulary: title, labels, horizontal x-axis and vertical y-axis, data series, tooltip, source and credits. image to come
Identify the chart type that best matches your story and data format.
Draw visual comparisons that are easy for readers to understand, rather than confusing them (adapted from Gourley p. 19). image to come
Do the math for your readers. Based on your data story, decide if you should show absolute numbers, percentages, or percent change (Wong pp. 23-25, 104-107). image and link to come
Order categories logically—either alphabetically, by value, or sequentially—depending on your data story (Gourley, p. 19; Wong pp. 70-71). image to come
For long labels, use horizontal bar charts instead of vertical column charts (Wong p. 66). image to come
On bar and column charts, start the vertical y-axis at zero, and choose natural increments (Wong pp. 51-52). But line charts do not need to start at zero, and can focus on specific ranges. See also the How to Lie with Charts and Maps chapter in this book. image and link to come
Beware of pie charts. Most readers cannot accurately estimate sizes of different slices. Consider other ways to show part-to-whole relationships, such as bar/column charts, or stacked bar/column charts (Few 2007, pp. 2-4; Wong p. 79). image to come
If you choose to use a pie chart, then show no more than 5 slices, and places the largest slices closest to the top at 12 o’clock (Wong, pp. 74-75). image to come
- Words matter as much as pictures.
- Add meaningful titles, labels, and annotations to draw attention to your data story.
- Keep typography simple, and use bold type sparingly to highlight your key insights (Wong p. 32; Knaflic pp. 107, 111).
- image to come
On static charts, label items directly when possible. (On interactive charts, designers may need to rely on tooltips and text.) Insert a legend in a logical place for readers (Wong, p. 56).
Add source credits and bylines—with links to view data tables and details—to build credibility and accountability.
Avoid “chart junk”–such as 3D perspective, shadows, and unnecessary ornaments—which distract readers from your data story. Never use 3D unless you are plotting three-dimensional data (Tufte p. to come, Wong p. 62, Knaflic p. 65). image to come
De-clutter charts (Knaflic pp. 91-98, 130-135).
- Choose colors wisely.
- Use color to logically organize your data. Avoid random colors (Wong pp. 40, 44).
- Avoid bad combinations from opposite sides of color wheel, such as red/green or yellow/blue (Wong pp. 40, 44).
- Use contrast (such as color vs gray) to call attention to your data story (Knaflic pp. 87-88)
- image to come
Stephanie D. H. Evergreen, Effective Data Visualization: The Right Chart for the Right Data, (Los Angeles: SAGE Publications, Inc, 2016)
Stephen Few, Now You See It: Simple Visualization Techniques for Quantitative Analysis, (Oakland, Calif: Analytics Press, 2009)
Stephen Few, “Save the Pies for Dessert [critique of pie charts],” Visual Business Intelligence Newsletter, 2007, 1–14, http://www.perceptualedge.com/articles/visual_business_intelligence/save_the_pies_for_dessert.pdf
Stephen Few, Show Me the Numbers: Designing Tables and Graphs to Enlighten, Second edition (Burlingame, CA: Analytics Press, 2012)
Drew Gourley, How to Use Data Visualization to Win Over Your Audience, (Visage and Hubspot, June 2015), https://visage.co/content/data-viz-win-audience
Cole Nussbaumer Knaflic, Storytelling with Data: A Data Visualization Guide for Business Professionals, (Hoboken, New Jersey: Wiley, 2015)
Cole Nussbaumer Knalfic, “An Updated Post on Pies,” StoryTelling with Data, February 16, 2017, http://www.storytellingwithdata.com/blog/2017/1/10/an-updated-post-on-pies
Wayne Lytle, Viz-O-Matic: The Dangers of Glitziness and Other Visualization Faux Pas, 1993 video shared on YouTube, https://www.youtube.com/watch?v=fP-7rhb-qMg
Isabel Meirelles, Design for Information: An Introduction to the Histories, Theories, and Best Practices Behind Effective Information Visualizations (Rockport Publishers, 2013), http://isabelmeirelles.com/book-design-for-information/
Tableau, Visual Analysis Best Practices: A Guidebook, n.d., http://www.tableau.com/sites/default/files/media/whitepaper_visual-analysis-guidebook_0.pdf.
Edward R. Tufte, Beautiful Evidence (Graphics Press, 2006)
“WTF Visualizations: Visualizations That Make No Sense,” 2017, http://viz.wtf.
xkcd, “University Website,” accessed February 12, 2017, https://xkcd.com/773/
Nathan Yau, “One Dataset, Visualized 25 Ways,” FlowingData, January 24, 2017, http://flowingdata.com/2017/01/24/one-dataset-visualized-25-ways/
Nathan Yau, “Best Data Visualization Projects of 2016,” FlowingData, December 29, 2016, http://flowingdata.com/2016/12/29/best-data-visualization-projects-of-2016/