Survey: Are Your Data Visualizations Negatively Impacting Your Company?


Data visualizations, when done right, can provide viewers with an immediate understanding of the implications and key takeaways of a data set. Be it a business report, a hurricane map or even a plot of words used in various Metallica albums, a data visualization can be a powerful method of communication.

When done poorly, however, data visualizations can have the opposite effect. They might confuse clients, embarrass you in front of your boss or damage your company’s reputation as a provider of accurate information. So, what effect are your data visualizations having on the people who see them?

To explore how people identify and respond to bad data, Software Advice created an online survey, collecting nearly 800 total responses from randomly selected U.S. adults. We also spoke to data visualization experts for advice on how you should be presenting your data. Here’s what we found.

Most People Can Spot Bad Data Visualizations

When it comes to making mistakes in your charts, graphs and other visual data representations, do people really notice? According to our survey, they do.

Respondents’ Ability to Identify Bad Data Visualization


For the first part of the survey, respondents were offered examples of data visualizations that were either presented poorly or presented well, and were then asked to identify which choices had something wrong with them. Two of the options were wrong, and more than half of the respondents were able to recognize this.


The first option was an image portraying percentages of “yes” and “no” answers, scaled with incorrect proportions. This faulty visualization was correctly identified as problematic by 22 percent of respondents. The third option was a pie chart with percentages that added up to more than 100 percent. This poorly-presented visualization was correctly identified as problematic by 34 percent of respondents.


Two of the other options were charts that had nothing wrong with them. Nonetheless, these were still identified as “wrong” by a combined 20 percent of respondents. The final choice, which was “none of these” (and was incorrect, as two of the options were, in fact, wrong), was selected by 24 percent of respondents.

Data Visualizations Respondents Identified as Wrong


In a nutshell, 56 percent of respondents were able to spot an incorrect data visualization on a whim—so it’s easy to imagine how much this percentage might increase if the data had to do with profit margins at respondents’ quarterly business meeting.

Bad Visualizations Cause Negative Opinions of the Presenter

We also asked respondents how likely they were to form a negative opinion of an employee or a company that presented them with inaccurate data visualizations. A combined total of over 60 percent of respondents said they were either “very,” “moderately” or “minimally” likely to form a negative opinion of the presenter based on an inaccurate data visualization.

Likelihood Bad Data Visualizations Will Cause Respondents to Form Negative Opinion


Only 9 percent of respondents said they were unlikely to form a negative opinion based on this, while more than 30 percent of respondents said they were unsure or didn’t care.

So if people are likely to spot mistakes in your data visualizations, and even more likely to form negative opinions of you based on those errors, what are some ways to avoid this trouble altogether? To find out, we spoke with data visualization experts across various industries. Here are the questions you should ask yourself to ensure that your data visualizations aren’t negatively impacting your company.

Is Your Data Presented in the Right Format?

The first way to guarantee that your data visualizations are as clear and concise as possible is to make sure you’re presenting them in formats that make sense, says Stefan Schmitz, the senior director of business intelligence (BI) Solutions Management for BI software vendor SAP Lumira.

“When you’re comparing the relative size of data values along nominal items, bar or column charts are a good choice,” Schmitz advises. “To understand how two variables relate to each other and to identify correlations, use a scatterplot or a bubble chart.”


Example of a bubble chart

Schmitz adds that an effective technique for visualizing changes in relationships is to animate a visualization through time.

“For example, in the case of a scatterplot, animation through time allows a viewer to see clusters forming and data points following certain paths in a two-dimensional space,” he says.


Example of a scatterplot

Rob Nelson, the founder of Grow, which provides BI dashboard visualizations for small and mid-size businesses, echoes Schmitz’s advice.

“The key is finding the visualization that works well for the metrics you want to manage,” he says. “Not everything is going to fit into a pie chart.”

Does Your Data Visualization Pass the ‘10-Foot Test?’

Once you’ve chosen the correct format with which to visualize your data, presenting it clearly and in an easy-to-understand way is still of the utmost importance.

“The whole goal of a data visualization is to bring the information forward in a manner that becomes instantly understandable,” says Kate Bagoy, the head of user experience and design for Alma, an Oregon-based company that provides data visualizations for students and educators.

If you think you’ve created a visualization that is clear and easy-to-understand, Bagoy recommends checking to see if it passes the “10-Foot Test.”

“If you put a data visualization on an 8.5- by 11-inch piece of paper and then step back 10 feet, can you still get a sense of what it is actually portraying?”

Bagoy explains that if you can, you’re likely on the right track. However, if you have to sit and analyze an infographic or other type of data visualization in-depth, then you’re missing the point.

“The entire point of a data visualization is the nearly instantaneous understanding of the point that you’re trying to communicate,” she says.


Example of a data visualization that would pass the “10-foot test”

Does Your Visualization Use Color Properly?

Another valuable tip for presenting clear and easy-to-understand information is the proper use of color within your visualization. Schmitz recommends using color to emphasize “only those components in a chart that actually represent values and items.

“Everything else that does not represent data should be just visible, but not distract from the data components,” he says. “Chart axes and check marks, if they’re needed at all, should be faint. Using a neutral light gray works best.”

Schmitz also recommends always rendering the background of a visualization a uniform color.

“A simple white works best in most cases,” he says. “Don’t use different fill colors for different parts of the surrounding area. And even if it looks nice, stay away from using a gradient background fill color. The gradient just distracts, and makes it a lot harder to compare the colors of the actual data components.”


Example of a bar chart with proper use of color. Note the uniform colors of the bars and the simple, white background.

From there, Schmitz recommends using different colors only when they correspond to differences of meaning in the data.

“For example, in a simple bar chart, where the labels on the axis already identify each bar, there’s no need to also use a unique color for each bar,” he says. “It’s a best practice to use natural colors, and [only use] brighter or darker colors to emphasize data that requires greater attention.”

Is Your Data Visualization Meaningful to Your Audience?

Finally, to ensure that your data visualizations have the impact you intended, make sure that what you’ve created has the proper context and meaning to appeal to your audience.

Mark Herschberg, who teaches courses involving data visualization at the Massachusetts Institute of Technology (MIT), says that oftentimes, the people creating data visualizations are BI professionals or data scientists who lack the business background of those who might be on the receiving end of their presentations.

“The data scientists usually have a math background, and are really good at understanding how they arrived at the data—but they might not understand the business as well as the sales guy or the marketing guy, or someone who has been in the industry for years,” Herschberg says.

“So the challenge here is that the data scientist shouldn’t just talk about, ‘this is a 72.3 percent downturn in this quarter, and we think it’s due to this seasonality,’” he adds. “They should talk about what it means to the business. They have to put things in terms that are meaningful to the actions of their audience.”

In other words: If you can provide context and meaning to your data visualizations, your audience will be more willing to listen.

“When you start out by saying, ‘this is the impact,’ you give people a preview of what the punchline is and why this is relevant to them—then, they’re going to pay much closer attention to the information,” he explains.


The clearest takeaway from this research is that recipients of data visualizations, be they colleagues or customers, are likely to scrutinize those representations and form opinions of the presenter based on their quality. Even though this was not true for a significant portion of the survey’s respondents, it was certainly true for the majority of them. Thus, a poorly-made data visualization could have a significant negative impact on your company.

Armed with that information, a key recommendation would be to put in the effort needed to ensure your data visualizations are not only clear and concise, but accurate and presented in a context that is meaningful and relevant to your audience.

Data presentations, when done correctly, are a valuable method of conveying information to others without wasting time or effort. Allowing someone else to see the data and understand it for themselves is often the best way to drive home the very point you want to express.


To find the data in this report, we conducted a three-day online survey with two questions, and gathered nearly 800 responses from randomly selected respondents within the United States. We worded the questions to ensure that each respondent fully understood their meaning and the topic at hand.

Sources attributed and products referenced in this article may or may not represent partner vendors of Software Advice, but vendor status is never used as a basis for selection. To further discuss this report, or obtain access to any of the charts above, feel free to contact me at

Bubble Chart” created by Albin Olsson used under GNU / cropped and resized.

Scatter diagram for quality characteristic created by Daniel Pinfield used under CC BY-SA 3.0 / cropped and resized.

Annual Arctic Sea Ice Minimum” created by NASA used under public domain / cropped and resized.

Share this post:  

About the Author

Abe Selig joined Software Advice in 2014. As the Managing Editor of Plotting Success, Abe analyzes and writes about BI trends and tools. He also writes content related to supply chain management.

Connect with Abe Selig via: 
Email  | Google+  | LinkedIn