Cole Nussbaumer Knaflic, industry thought leader and author of “Storytelling with Data,” an Amazon bestseller, recently sat down with iCharts to share her expertise. In this series of blog posts, we’re bringing you her insights on data visualization best practices.

In the last post, Knaflic shared her thoughts on the importance of understanding your audience when creating data visualizations. Here, she provides her perspective on the do’s and don’ts of effective data visualization.

Do: Put the audience first. “This means understanding how your audience sees and how you can leverage visual cues to make your data visualizations accessible and understandable to your audience,” Knaflic says.

Do: Declutter. For Knaflic, creating an effective data visualization means “leveraging white space and not overly cluttering your visuals and making them harder than they need to be for your audience to digest.”

Do: Practice. “Practice is definitely the way to get good at data visualization. Take some data and play with it and look at it in different ways. Put it into different types of graphs and see what it looks like with this color and that color. You can develop a personal data visualization style along the way.”

Do: Put thought into your use of color. According to Knaflic, “color used sparingly and strategically is one of your most powerful tools for signaling importance to your audience and drawing their attention where you want it.”

Do: Use text in your visualizations. “Sometimes when we’re communicating with data, it feels like words have no place with all the numbers and pictures. But words have a very important place in making those numbers and pictures accessible to the audience,” Knaflic explains. “You want to use text to label, to title, to explain. Don’t ever assume that somebody else looking at the same data visualization is going to walk away with the same conclusion. If there’s a conclusion that you want your audience to reach, put it in words.”

Don’t: Fall into the trap of data for the sake of data—according to Knaflic, there’s not a lot of meaning in doing so. Instead, think about where you want to drive action, and how you can use data to convince people of that. As Knaflic says, “One way to help make data understandable and relatable is to visualize it. By visualizing data and helping people see it and understand it, we can help incite understanding, incite action, and really drive change.”

Don’t: Use pie charts. Although Knaflic says she recognizes that intelligent people will disagree with her, she absolutely recommends avoiding pie charts. “But worse than pie charts is 3-D. Any sort of 3-D rendering of data visualization, just please don’t. The only exception is if you’re actually plotting a third dimension, and even there, it gets really tricky, really quickly,” Knaflic says. “And never use the 3-D exploding pie chart. It makes the data harder for the audience to get at, and there’s no good reason to do that. Just because something exists in a tool does not mean you should use it.”

Both “Storytelling with Data” (the book) and Storytelling with Data (the website) are packed with useful information on data visualization best practices. We also invite you to watch a three-minute video that demonstrates just how the visual analytics solution from iCharts helps companies extract more value from NetSuite.

Watch the full interview below.