We’ve reviewed many dataviz books
here at ZingChart, and are pleased to share a new review of Storytelling with Data by Cole Nussbaumer Knaflic. You might think another data visualization guide for business professionals is not necessary. But we found a lot to like in this book.
A Dataviz Book That Starts Out on the Right Note
The foreword begins with an Edward Tufte quote that we should all know:
“Power Corrupts. PowerPoint Corrupts Absolutely.”
This sets the tone for the level of dataviz that will be discussed in Storytelling with Data. There are no complex regressions or heavy duty calculations here. We’re assured this is information we can use in the type of charts we use everyday.
Just as we're about to settle into the book and get down to dataviz business, we see the acknowledgements. Awww.
A data visualization of the times in the author’s life when those she thanks have influenced her.
And then we hear from Cole Nussbaumer herself in the introduction. There’s a section titled, “We aren’t naturally good at storytelling with data.” This section was very unique to see in a dataviz book.
The author does not make us feel inadequate for our lack of innate dataviz expertise. She relates that most academic courses center either around stories or numbers, not stories about numbers. This brings the point of her book into focus. We will be learning lessons to move beyond showing data to storytelling with data.
And then she gets to the point and starts showing us before and after examples of data sets that simply show data, and those that tell a story.
Yes- that is an example to take in right away on page four.
Explanatory Analysis vs. Exploratory
In chapter one, Nussbaumer makes an important distinction between the two ways we work with data. We use exploratory analysis to look at data before we ever show a chart to someone else. When we’re ready to share some highlights, we use explanatory analysis.
The author says it is important to perform this second type of analysis, instead of showing our audience all the background work. She wants to let the audience concentrate on pearls of wisdom, instead of all the “oysters” it took to find them.
One great tactic Nussbaumer shares for being explanatory is to simply use text. This is a great point that is often overlooked when we have so many chart types at our disposal. Take a look at this example:
She calls this a text makeover. We can’t wait to use this in more visualizations.
A Few Recommendations from this Dataviz Book
If you’re telling your data’s story with a bar or column chart, Nussbaumer has a recommendation on how thick the bar should be. “Bars should be wider than the white space between them.”
However, bars should not be so wide that users compare their areas instead of lengths. Good to know!
Have you ever heard the term “dessert visual?” Neither did we. Apparently it refers to donut and pie charts. And Nussbaumer places these two charts in the “avoid” category for dataviz. She says donut charts are just as confusing for users to interpret as pie charts… if not more.
Speaking of pie charts, the author compares area charts to them in an interesting way. If you’re going to be critical of a pie chart because it makes comparing the areas of a slice difficult for users, area charts should be considered even more challenging.
You’re not only comparing areas of slices, you’re comparing completely different shapes. We’ll keep that in mind.
According to Nussbaumer, another term for clutter is “cognitive load.” This term allows you to include (or exclude) a lot more from your chart than simply elements such as grid lines. According to the author, this can cover more than “stuff,” including:
Hierarchy of items in the chart area
Alignment of items
Quick Chart Clean Up
This tidbit is a great: one of Nussbaumer’s biggest dataviz pet peeves is trailing zeros on y-axis labels. “They carry no informative value, and yet make the numbers look more complicated than they are!”
Cleaning up the trailing zeros on your y-axis is a quick and easy dataviz win.
Use of Color
Nussbaumer mentions a color guideline we have not seen in other dataviz books: non-strategic use of color and contrast. She states that taking in information with random colors, especially without contrast between the color values, is time consuming. Her example in this case is powerful, even at a quick glance:
Other Color Considerations
Colorblindness is another color consideration discussed in the book that you don’t often see in other dataviz books. Nussbaumer shares several resources to use on the web or your local machine to analyze color choices to ensure they are visible and discernable to colorblind users.
Color in culture is another unique issues the author introduces. Besides thinking about the tone colors conveys (red is negative), she mentions David McCandless’s studies on visualizations in different cultures. He explored how users interpreted the meaning of visualizations when different colors were used.
Getting People On Board
We also really liked the section on acceptance in chapter five, Think Like a Designer. Nussbaumer wrote, “For a design to be effective, it must be accepted by its intended audience.” This can be especially tricky when the audience is your boss. So she shares a few strategies to help get people on board with storytelling with data, instead of just showing lots of charts, or doing things the way they’ve always been done. The broad strokes include:
Articulate the benefits of this approach - mention some improved observations users have made with this new outlook
Show the side-by-side - Sometimes a clear before and after can help make the shift
Provide multiple options - There are many ways to tell stories with data, so offering a few options can keep people feeling involved in the process
Get a vocal member on board - Finding a few key influencers to vocally support your new dataviz can help others follow
More Dataviz Book Reviews
Storytelling with Data was a fun dataviz book to review. It contained many points we look forward to using in our upcoming chart and graph projects. We look forward to more books by Cole Nussbaumer Knaflic. Until then, which book should we read next? Share your favorite dataviz book with us in the comments section below.