Common Data Visualization Mistakes You Can Fix Right Now

Data visualization aims to take a spreadsheet with hundreds of columns and numbers and make sense of it.

If you're looking at a spreadsheet, the odds of you being able to see outliers or trends quickly are pretty small. Data visualization is how we make sense of the numbers and communicate those insights to those not looking at the spreadsheets.

But not all data visualization is made equal. There are excellent strategies and not-so-excellent ones. Better data visualizations can communicate the story of the data quickly, accurately and with impact.

Most of us know significant data errors like being unable to see a trend because the Y-axis starts at 100 rather than 0 or using the wrong chart type. But there are also subtle mistakes that can be made in data visualization, and understanding how to avoid them is critical.

In this article, you'll learn some simple data visualization mistakes and how to avoid them to be a better data storyteller.

?

1. Over-utilizing Pie Charts

Oh, pie. Where do we begin? Simply put, pie charts are problematic. Yes, there is a time and place for a pie chart. Usually, they'll do a good job if you need to show a simple comparison between a few data points. But rarely is it ever executed well.

?

Here are just a few problems with pie charts:

1. Difficult to compare values: Pie charts can make it difficult to accurately compare the sizes of different segments, mainly if there are many or small segments.
2. Limited data representation: Pie charts can only represent a single data series, which makes them less helpful in showing multiple data sets or relationships between variables.
3. Labeling issues: When there are small segments in a pie chart, it can be challenging to label them accurately, and the labels can become crowded and difficult to read.?
4. Overemphasis on part-to-whole relationships: Pie charts focus on the relationship between individual segments and the whole, which can sometimes overshadow other important data trends or relationships.
5. Misleading with 3D effects: 3D pie charts can distort the data, making it difficult to interpret the relative sizes of the segments accurately.
6. Color dependence: Accurate interpretation of pie charts often relies on color, which can be problematic for individuals with color vision deficiencies or when printing in grayscale.
7. Biased perception: The human brain has a hard time perceiving angles and areas, which can lead to biased interpretations of the data in a pie chart.

?

For fun, we made this pie chart explaining the problem with pie charts.

?

2. Sticking With Default Settings

You might explore data for hours or days (maybe weeks!). The defaults definitely help you do some exploratory diving into the spreadsheets. You're making sense of it for yourself.

But one big mistake with data visualization is just going with the defaults in your settings. You input the data, click the chart button, and whatever shows up is the final draft. The problem is that most data visualization software tries to showcase all the tool's features with the default setting.

They'll add in every color available or all of the grid lines, numbers, axis and labels. Instead of letting the defaults decide what your stakeholders pay attention to, you should be more intentional about your design.

Take ownership of the visualization process by strategically choosing your colors, labels and axis. The best strategy is to remove any distractions like data that doesn't matter, colors that don't serve a purpose or text that doesn't add context.

We do this to remove the noise because it will add confusion and require the audience to do more work to understand your chart.

One idea to reduce noise is to eliminate the chart legend. The problem with the chart legend is that your audience has to look at the chart legend to understand what color or section is relevant. And then look back at the chart, and back at the legend and so on and so on.

You can eliminate this back and forth by incorporating the chart legend into the chart so the audience can focus on the story you're trying to tell.

?

3. Forgetting the Audience

You shouldn't create a bunch of charts and assume your audience will understand them the same way you do. You have context and experience with the database you used to create the chart.

Instead, you should consider your audience and the context they'll need to make the same conclusions and understand your insights.

For example, maybe you think a box chart is exciting and explains your insights perfectly. If your audience doesn't know how to read a box chart, you might spend more time explaining the chart than providing insights.

The best strategy is to think of the chart that will be most intuitive for your audience. Your audience should be able to look at a chart, understand the story it's telling, and draw conclusions quickly without a lot of extra explanation.

?

4. Big Fonts Aren't Data Visualization

Some people will take one number and make it into a big font, like $4 million—that's our revenue for the year. This one number; if you bold it, center it and make the font really big, it will capture attention—no doubt. But it doesn't give any context, and that's what you want to provide.

The problem with this BIG number (maybe next to an icon or image) in the center of a page is not data visualization.

The reason it's not data visualization is because data visualization needs a benchmark to tell a story. For some people, $4 million might be exciting; for others, it could be disastrous. Communicating with data requires you to maintain control over the narrative.

Whenever you use just one number without context, you're handing over control to people's individual experiences.

Instead, you might need to use that number and compare it to competitors in the market or historical trends. Is $4 million positive, negative or neutral? Is the company on track or off track?

The number by itself leaves everyone thinking, "so what?" We have to answer that question.

You might be thinking, "But what about percentages? The context is the comparison to 100%"

However, even with percentages, you still encounter the "so what?" question. How does the percentage compare with the last quarter or last year? How does it compare with other product lines or maybe your competitors?

?

5. Not Adding Variety To Your Visualizations

The big three are line charts, bar charts or pie graphs. And those are highly effective when designed well. The problem is whenever we see the same chart repeatedly, the audience might forget what chart belonged to what insight—everything gets a little ... blurry.

Imagine sitting through a presentation with 50 slides. There might be bar chart after bar chart (or worse, pie chart after pie chart).

Mixing it up and providing variety keeps the audience engaged. Not only does this break up the monotony of a presentation, but it also helps reinforce critical insights.

So don't be afraid to reach for those other data visualization tools in your toolbox. A word cloud might help you emphasize keywords or phrases from market visits. A scatter plot could help you demonstrate correlations between two variables like NPS score and retention. An area chart might help product managers visualize user adoption over time.

The goal is to keep the story flowing and maintain your audience's attention. There's no limit to the number of available tools and techniques when it comes to data visualization.

If you're looking for some data visualization inspiration and best practices, check out our latest ebook: Put Down The Pie (Chart): A Comprehensive Guide to Data Visualization for Product Managers

?

The Goal: Become A Better Data Storyteller

Why do we want to avoid these mistakes? Simple, all want to be better data storytellers. It gives credibility to our strategies and helps earn buy-in at all levels of the organization.

Data visualization isn't just about slapping a few numbers together in Excel or Google Sheets and calling it a day. It requires thoughtful consideration of your audience, the story you're trying to tell and how it will be remembered.

?

Next Steps: Enroll in Pragmatic's Insight Course

Ready to take your data analysis skills to the next level and transform your product strategies? Enroll now in Pragmatic's Insight course. Our expert instructors will teach you practical techniques to identify patterns within data, prioritize problems, and develop a scalable process for successful data projects. With our hands-on approach, you'll gain the skills to turn data into powerful insights and make informed decisions that will take your products to new heights.

LEARN MORE