29.03.2024

6 Mistakes Ruining Your Charts and Infographics

A word about words: When Scott says “chart” or “graph” or “data visualization” or “dataviz” or “information visualization” or “infoviz” or “information graphic” or “infographic,” he refers broadly to the visual communication of data. I’m using the terms “chart” and “infographics” in this same broad way.

Want some tips guaranteed to result in bad charts?

Of course you don’t. Yet sometimes we learn best from things gone wrong. That’s why, in this article, I offer some of Content Marketing World speaker Scott Berinato’s advice flipped on its head.

Scott shared great tips on how to get data visualization right in his talk, Data Visualization and Creating Good Charts. I’m pointing out how, if you’re not following his advice, you’re surely confusing, boring, and bothering your audiences.

You tell without showing

Let’s start with the ultimate bad chart: no chart where one is needed.

Take a note from this fictitious example from Scott. Which version of these directions (left or right) do you prefer? Too easy, I know. If you’re creating directions with text only, you’re guaranteeing maximum inscrutability.

fire-escape-plan-chart

The same problem arises when you describe the significance of numbers with a wall of words. Check out how the paragraphs on the left force you to dig to discover whose fortunes went up in 2015 – something that the chart makes obvious.

fortunes-chart-example

Omitting images wallops people with the worst possible information experience. Don’t make words do all the work; showing is more powerful than telling when it comes to directions and data trends.

You include extraneous details

Details go a long way toward making charts accurate and credible. Yet one of the easiest ways to ruin a good chart is to pack in more details than people need or could be expected to care about.

For example, a colleague of Scott’s sent him this photo from a presentation she was watching. Here’s Scott critique:

Don’t be this guy. What is he trying to convey? To whom? I guarantee you he’s not talking about ideas. What is he talking about? Well the x-axis is this, and the red bars mean this … He’s talking about the mess he has on his screen, and he’s not getting any ideas across. You know what the audience is doing? Tuning out. They’re texting pictures to their friends and saying, ‘Get a load of this guy.’

chart-extraneous-detail-example

If you’re rolling out charts with extra gridlines, extra labels along the axes, and extra tick marks on the bottom, you’ll have people scrambling for their magnifying glasses. Compare the charts below. Which do you find easier to follow?

chart-extra-detail-example

You choose inappropriate chart types

Many tools automatically convert data to charts. Anyone can turn a spreadsheet into a chart with a few clicks. But choosing the wrong chart type makes viewers scrunch their eyebrows in confusion.

These two examples give a glimpse of the results of poor data visualization choices.

In this first chart, if you squint long enough over the green line, you figure out that this data has nothing to do with a trend. “You might say that’s a ridiculous example,” Scott says, “but I see this all the time in board presentations.”

faux-trend-line-example

This faux trend line is a common example of an inappropriate chart type. If a bar chart had been used, readers would instantly get that it is comparing six departments’ levels of travel spending.

In the chart below, which is not related to the previous one, notice the 3D bars and the jut of the graph itself.

3d-chart-example

Scott, who serves as senior editor of Harvard Business Review and wrote the book Good Charts: The HBR Guide to Making Smarter, More Persuasive Data Visualizations, considers 3D charts universally inappropriate: “Don’t bend things. Don’t make charts 3D. It’s no good. I don’t know why software still lets us do this. It’s silly.”

Scott’s opinion echoes that of dataviz authority Edward Tufte, a famous denouncer of 3D charts who coined the term “chartjunk.” Chartjunk often includes “graphical decoration,” writes the Yale professor emeritus in his book The Visual Display of Quantitative Information. He’s talking about “ink that does not tell the viewer anything new … non-data-ink or redundant data-ink.”

We’re looking at you, little gray squares under the blue bars.

Unfortunately, graphical decoration, as Edward writes, “comes cheaper than the hard work required to produce intriguing numbers and secure evidence.” This man isn’t called “the da Vinci of data” and “the Galileo of graphics” for nothing.

Bad charts were never so easy to make. Avoid the temptation of silly software features.

You use confusing x- and y-axis values

A surefire way to keep your audience guessing about what story your chart is telling is to choose your x- and y-axis values carelessly.

Here’s a before-and-after example showing how the y-axis can obscure the story. At first glance, you might not see much difference between the chart on the left and the one on the right.

The chart on the left shows gold and silver prices in dollars with two y-axes – one on either side – which muddies comparisons.

The chart on the right shows the percentage of change in the prices. Silver (the blue line) has higher highs and lower lows than gold (the orange line). In this chart, Scott says, viewers quickly see that silver is a more volatile investment than gold.

price-gold-silver-chart

To help people quickly see the stories told in your charts, stick with y-axes that don’t force people to dig for what they want to know.

Make sure your x- and y-axis data tells the story you want to tell. Here’s one of Scott’s favorite examples from his own experience. (He has changed the details to keep the data confidential.) He was getting ready to present a chart like the one on the left at a board meeting. The yellow bars were supposed to show customer purchasing levels (y-axis) throughout the day (x-axis).

customer-purchase-activity-chart

In fact, the chart on the left tells the wrong story, although you’d never know it. Scott’s colleague discovered that the time of day along the x-axis indicated the time the purchase was recorded on the server in New York – not the time of purchase in the customer’s time zone. The chart on the right shows the adjusted data to reflect the customer’s time of purchase.

Give these two charts the squint test. You can’t help but notice the roller-coaster dip during the wee hours in the updated chart. It practically shouts, “Look what happens to spending when people are sleeping.”

You zoom to frame the data that supports your view

Mark Twain popularized the line, “There are three kinds of lies – lies, damn lies, and statistics.” Bad data visualizations lie without exactly lying.

Consider these two charts. Both show accurate data about sales of vinyl records. The chart on the left zooms in on sales between 1993 and 2014, showing a dramatic surge. The chart on the right zooms out to include sales since 1973. The “surge” in the 21st century all but disappears in that context.

vinyl-sales-chart

Shifting the data frame confuses the story.

You use only static images

If you use only static images to tell stories with numbers, you may be missing out on one of the most compelling ways to convey data: video (including animation). In an interview with Clare McDermott, for example, Scott points to this video, which depicts World War II death tallies in a way that no static charts alone could:

The Fallen of World War II from Neil Halloran on Vimeo.

Per Clare’s article, Scott points to three things that make this video so impactful:

  1. The author turned data into a story with a setup, conflict, and resolution.
  2. He used animation to do what animation does best: show change.
  3. He kept things simple by using only three chart types: stacked bars, unit charts, and stacked area charts.

If your numbers tell a powerful story involving change, consider using moving images – even simple ones. Think beyond static charts.

Conclusion

Making bad charts takes hardly any thought. Far too many people do it. Here’s a rundown of the data visualization mistakes I’ve covered:

  • Telling without showing
  • Packing in extraneous details
  • Choosing inappropriate chart types
  • Using the wrong x- and y-axis values
  • Zooming in (or out) to frame the data that supports your view
  • Sticking with static images only

If you’re making these mistakes, you’re likely misleading, befuddling, and boring your audience. Avoid these pitfalls, as Scott suggests, to make charts that clarify – and maybe even inspire.

What are your favorite techniques for creating charts that work?

Leave a Reply

Your email address will not be published. Required fields are marked *