2 upvotes 1 discussion

You turn on the projector, giving way to a mess of graphs, charts and numbers.

“That looked better on paper," you think to yourself. "Wait, no, it looked terrible there, too,” You look around, hoping nobody noticed your faulty work. The perplexed looks on your colleagues' faces confirm your worst fear: You screwed up. This meeting was doomed before it even started.

Had you been equipped with the data visualizations knowledge you needed, this mishap could have gone a much better way.

Why do we need data visualization?


When I see raw data on a spreadsheet, my head starts to spin. It’s hard to grasp what that data is really saying. Computers read numbers as a language. People, well, not so much. We are visual beings.


Data visualization is part science, part art, and part language. It’s science in that you need to use a systematic approach to organize your data. It’s art in that you use colors, shapes, and textures -- or a lack thereof -- to communicate your data in a clear and compelling way. And, most importantly, data visualization is a language. It translates complex concepts into key findings everyone can understand.


To hit the trifecta just described, you need to set yourself up for success. Follow these do's and don'ts the next time a data visualization project lands on your desk:

Do: Tailor your visualization

To start off on the right foot, you need to envision the end -- the end goal, that is. Ask the good ole, who, what, when, where, why, how, and a bonus one: How much?

This will vary based not only on the data itself but also who you're presenting it to. As the golden rule of marketing commands: Know your audience.

Ask yourself who benefits the most from the data visualization -- Who's it for exactly? Your sales team? Your boss? Your boss's boss? 

Don't: Forget core design principles

Tailoring your data visualization is one thing, but reconstructing it is another. Good data visualizations strike a balance between readability and practicality, making them both enjoyable and beneficial to consume. Don't think catering to your CMO's love of polka dots will score you any points.

One of the strongest proponents of goal-driven data is statistician and artist Edward Tufte, who's "been preaching the merits of quality data visualization since before the world wide web."  If data visualization is something you want to pursue, Tufte's works are a must-read.

When in doubt about data aesthetics, refer back to the foundations of good visual design (i.e. your rules to live by). Key points to remember include:

  • Use color theory to draw attention to certain pieces of data and differentiate one variable from another. Think of color as an accent, made to drive your point home—the exclamation point on a piece of data.
  • Pick a classic typeface that's legible and professional. Basics like Arial, Helvetica, and Georgia are always a good pick. If you company has a style guide, refer back to it to find out if any specific typefaces are used.
  • Avoid clutter at all costs, and embrace white space around or between data visualizations. Strong designers know when to stop designing and let the work speak for itself.

Do: Choose the right visual format

Avoid defaulting to PowerPoints’ "one click ta-da! Chart! Done." Give your data the attention it deserves. The wrong format choice could very well communicate the wrong message.

How wrong? Let's just agree you won't take that risk.

Harvard Business Review gives a great example of this in an article outlining important data viz questions to ask. They include a side-by-side comparison of a pie chart and a bar chart, both displaying information about investment amounts.

Looking at the pie chart gives us an understanding of how investment funds were divided. We see using this general overview that most investment money went to international and large-cap U.S. stocks. The bar chart, however, provides a more detailed breakdown of how each investment scales in comparison to one another. This chart clearly gives us the same findings as did the pie chart but with more granularity.

As HBR explains, "When you choose how to visualize your data, you’re deciding what type of relationship you want to emphasize."

For reference, here's a list of the most common data visualization formats, along with descriptions of when to use them:


Charts and plots

  • Bar charts are good for comparing nominal, like the number of red shirts sold per quarter, or ordinal data
  • Stacked bar charts 
  • Bubble charts are useful for showing different scales, like XYZ.
  • Line charts are useful when showing change over time, like site traffic throughout the year.
  • Pie charts can be used to compare percentages of a whole between 2 or 3 items.
  • Sparklines can be great for quickly displaying important variations, such as business trends.


Short for "information graphic," an infographic is another effective tool for presenting data. Infographics emphasize creativity, shareability, and storytelling to communicate data in an engaging way. The most common types of infographics include:

  • Timelines
  • By-the-numbers
  • Anatomies
  • Geographical maps
  • Venn diagrams

Opt for an infographic only if you have the creative resources to tackle it. What that really comes down to is access to a design program like Photoshop or Illustrator and someone who knows how to use them... well.

Don't: Include more than you need

It can be tempting to throw tons of stats at your coworkers or boss, especially when you’re proposing a new idea or hoping for more resources. That being said, more is not always better

While you want to highlight all of your pertinent findings, you also want to keep your data visuals streamlined and simple. In order to do that, you should abide by the following rules:

  • Don't add pointless design elements like slide transitions, patterned backgrounds, or multicolored text. These are distracting at worst and off-putting at best.
  • Don't write lengthy explanations about your data findings. The takeaway should speak for itself, and you should know the data well enough to elaborate if asked.
  • Don't use technical language or industry jargon in titles and captions. Your audience shouldn't need a dictionary to decipher what you say.

Do: Label your visualization

Your graph is only as good as its labels -- Otherwise it's just lines on a screen.


Useful labels include titles for charts and axes, a key or legend to differentiate visuals, and short pull-out quotes or descriptions to clarify important points.


It's also good to keep in mind that you may have colorblind readers, so avoid using red and green -- the most commonly unseen colors -- unless absolutely necessary.



Moving forward: Data visualizations can be an extremely effective way to convey important data in a succinct way. Just remember to keep things simple and tidy, and pick the right data and tools for the job. 

What data viz principle do you find the hardest to implement?