Good Charts Workbook: Tips, Tools, and Exercises for Making Better Data Visualizations

by Scott Berinato

A good visual representation of data can be invaluable for communicating the meaning behind the numbers. This book walks the reader through the thought process and choices in creating visualizations for a variety of cases. “It’s rare you don’t have to make a trade-off to create a good chart… Most of the time there isn’t one right answer, one right chart.” The main topics covered in the book are clarity, color, chart types, and persuasion.

Clarity. Berinato emphasizes decluttering to put the focus on what you want to communicate. “Take stuff away… Remove redundancy… Limit color and eye travel… Use your headline to describe the main idea of a chart, not its structure.”

“If you find yourself using text captions to explain parts of your visual… then the chart isn’t clear. A clear chart communicates its ideas with little or no intervention. It stands on its own.”

“In general, simplicity leads to clarity, but sometimes less is less. Simplicity fails when we’re forced to stop and think about what we’re looking at. If the audience is asking for information that’s not there, the chart is probably too simple.”

“Stick to conventions. Just as with statistical charts, heuristics matter. Time generally still moves to the right. Red means hot or danger. Green is good or safe. Hierarchies go from top to bottom. Our minds are so familiar with these ideas that it’s disruptive to change them.”

Color. “If you have time to focus on improving only one thing in your charts, go after color.  Most software can’t intuit a good use of color for your context. It can’t know how you want to group variables… Thus software tends to give every variable its own, somewhat randomly assigned color… Most charts start with too much color… Rainbows are pretty, but in charts they’re usually a detriment. Keeping track of what each color represents is difficult, and all the colors beg equally for attention.”

“Text, labels, and other marks that aren’t part of the marks that are conveying the data information are best left black or gray.”

One critique I have is the misuse of color terminology:

  • “When variables are inherently similar, use similar or complementary colors. When they are in opposition, use contrasting colors.” While the author’s meaning can be discerned from the context, the wording is contradictory. In art and design terminology, complementary colors are contrasting: red and green; blue and orange; yellow and violet; etc.
  • “Since the top two activities are types of leisure, two hues of the same color will show that they’re complementary, not contrasting. Passive leisure gets a lighter hue because it feels softer…” What he means is dark and light values of the same hue. However, in the context of data visualization, if the art term value could be confused with numerical values, perhaps dark and light shades would be a better word choice. And again, the same color cannot be its own complement.

Presentations. “One of the most common dataviz mistakes I see in presentations is when presenters try to stuff as many ideas as possible into one chart—or multiple charts on one slide—to keep the slide count down. I’d rather use two slides, each for a single idea or chart, than try to pack in the ideas.”

“Instead of piling new information on top of old, I do something subtly different. At each step new information is introduced and old information is removed. This forces focus. Only what I want to talk about it there for the audience to think about.”

“My principle is simple: I’d rather use 10 slides that each take 10 seconds to present than two slides that require five minutes to unpack. When you shift the unit of measure from a slide to an idea, and when you use multiple slides to build an idea, people stop thinking about how many pages you’re going through, because they’re utterly engaged.”

But context matters. An effective visual for a live presentation may be different than one for a printed document. “In a live presentation I have seconds to get people’s attention and help them understand. I can’t afford to present dense slides with many charts that require explanation. But if the viewers are reading on their own, on a screen or on paper, the sparse presentation slides may not be enough of a guide for them. They can control the pace, so I’m afforded the luxury of layering more detail into one space.”

Chart Types. “Most dataviz challenges can be handled by three chart types and their variants: line charts (stacked area, slope graph); bar charts (stacked bar, dot plot); and scatter plots (bubble chart, histogram).”

Appendix C helps the reader choose an appropriate chart type based on keywords. A bar chart or line graph might be the best choice to show a trend. You might select a pie chart or stacked area to show proportions. A glossary of the following chart types is included in Appendix A: 2×2 matrix, Alluvial diagram, Bar chart, Bubble chart, Bump chart, Dot plot, Flow chart, Geographical chart, Hierarchical chart, Histogram, Line chart, Lollipop chart, Metaphorical chart, Network diagram, Pie chart, Sankey diagram, Scatter plot, Slope chart, Small multiples, Stacked area chart, Stacked bar chart, Table, Treemap, and Unit chart.

Of course, not all graphics impart meaning. “At HBR we call those process diagrams that show cycles and other neat but meaningless procedures ‘crap circles,’ a term coined by senior editor Gardiner Morse.” That’s brilliant!

The author mentions a number of tools he has used, including Excel,, Sketches, Adobe Illustrator, Raw, Tableau, and R.  He also points out, “tools such as Coblis and Color Oracle make it easier than ever to see how you charts will look to those with some form of color blindness.”

This workbook is a follow-up to the author’s previous book, Good Charts: The HBR Guide to Making Smarter, More Persuasive Data Visualizations.

Berinato, Scott. Good Charts Workbook: Tips, Tools, and Exercises for Making Better Data Visualizations. Boston: Harvard Business Review Press, 2019. Buy from

Disclosure: As an Amazon Associate I earn from qualifying purchases. I received a review copy of this book.