5 Tips for Optimizing Charts

Charts can be incredibly challenging. They represent where the right brain and left brain meet… where computations and numbers collide with art and color. To be able to convey true meaning with graphics is a very special skill; here are 5 tips to help you along the way.

Line chart showing two years of chocolate pie revenue

1. Determine Your Message

Two important questions to ask yourself before you dive in to chart creation:

  • Who is your audience?
  • What is your message for them?

Cognitive Overload: Be Kind to Your Audience

One mistake a lot of us make is overloading the audience with too much information. This concept is called cognitive overload. Sometimes we zoom out in an attempt to show lots of data… and end up burying the story we are trying to tell in the process. The audience loses patience, and the entire message is lost.

Horribly messy pie chart titled Star Trek Fans

Consider this unfortunate pie chart… what story was the creator trying to tell? We are going to find some better solutions for this data shortly.

Audience Motivations

Some other important questions to ask are:

  • What is important to your audience?
  • What motivates them?
  • What is their level of knowledge on the topic?

All of these questions should influence the way you determine your message. We are going to revisit this first point frequently throughout this article. Even though it seems like the simplest of concepts, it is often the most forgotten.

2. Select the Right Chart Type

If the picture above gave you minor palpitations, part of the problem may be that someone picked the wrong chart to convey this information. There was a lot of data to comprehend here, more than a pie chart could feasibly tolerate. The chart below uses the exact same data set formatted as a column chart. Is this easier to understand?

Column chart showing star trek fans by state

Alright, the message is still pretty unclear, but we are moving in the right direction.

Here are a few general rules for selecting chart types:

Pie Charts

  • Pie charts should contain no more than 5 pieces, and preferably fewer than that.
  • Pie charts are supposed to show portions of a whole, so the whole should ideally be represented, even if you incorporate grouping (discussed next).
  • Remember your message and your audience (Tip #1), and then consider: are all your pie pieces all the same size? If so, is this pie chart really showing what you want to tell? Maybe it is… maybe your story is that all things are equally represented. But if that is not your message, consider a bar or column chart.
  • Consider incorporating data labels and callouts for further clarification.
Pie Chart titled Favorite Weekend day, showing 86% Saturday and 14% Sunday (fictional data).

Bar Charts and Column Charts

Bar and Column Charts are a great option if you have more information to display.

  • The main difference between bar charts and column charts is that bar charts are composed of horizontal data bars and column charts are composed of vertical data bars.
  • Both are great options for showing larger numbers of data sets.
  • Still, beware of cognitive overload with too much information (see Narrow your Focus for more tips below).
  • Clustered column charts are great for showing quick comparisons between small groups.
Clustered column chart showing dessert sales by three people.

Line Charts

  • Line charts do well comparing two competing data sets over a time period, like monthly sales figures this year stacked up against sales figures last year.
  • They also can help a viewer quickly ascertain overall trends at a glance.
  • One tip: when possible, start the Y axis at 0. This should already be the default setting.
Chocolate pie revenue line chart showing two years of data.

So Many More…

There are so many more charts to choose from! Check out this helpful Chart Chooser for more assistance with selecting the right type of chart.

3. Narrow Your Focus

For this section, I would like to revisit our original problem chart from the beginning. Remember, we started with this:

Horribly messy pie chart titled Star Trek Fans

And changed the chart type to a column chart:

Column chart showing star trek fans by state

…but this is still not an ideal situation.

What story do you want to tell?

Tip #1 will help you narrow your focus for this step. In this case, we had a (fictional) chart of total U.S. Star Trek fans broken down by state. Let’s explore a few stories you may want to tell with this data.

Avoid displaying too much information

Because all the states are represented, this means there are 50 data points. Do we really need to show all of this information? Perhaps, you decide that the story you want to tell is to show the states with the highest rates of Star Trek Fans.

How about removing the states that are not pertinent to your story? Let’s try to narrow our focus by filtering out some unnecessary information (i.e. the lower figures in the data set) with the filter button to the right of the chart.

Filter dropdown next to chart

This leaves us with a more manageable data set, down from 50 points to 7…. let’s keep going.

Column chart titled states with the most star trek fans, showing 7 states.

Highlight Important Information

Another useful technique to drive home a point is to use contrast to highlight important information, visually pulling it to the foreground. Select any data bar, right click, and select Fill to choose any color in the color wheel.

Right click menu, fill selected.

Let’s make use of this opportunity to pull the highest states to the foreground with a deep color, and grey out the others, pushing them to the background.

Take a look at the same filtered graph, recolored. Maybe the story you want to tell is that Hawaii, Kansas and New York had the highest populations of Star Trek fans… if so, this could be a cool way to do it.

Column chart for states with the most star trek fans, with highest figures in a darker red shade.

Group information together

We initially started with a pie chart, and there still might be a case for this type of chart with this data. Grouping together pieces of information is another great way to focus in on your overall message.

Perhaps your goal is to spotlight Kansas, and tell the audience that Kansas contains 5% of the country’s Star Trek Fans. Why not group together the states that are less relevant? We can also incorporate the previous highlighting technique.

This is much easier on the eyes than the original pie chart.

Star trek fans by state, showing Kansas pulled out on its own and the other states grouped together.

Do you see why asking yourself the questions in Tip #1 can put you on the right track for creating a meaningful chart?

4. Choose Words Carefully

Take a look at the chart below. ..

Very wordy pie chart

How long did it take you to grasp the meaning of this chart?

Wherever possible, keep the words to a minimum. And when in doubt, refer back to Tip #1 and ask yourself:

  • Who is your audience?
  • What is your message?
  • How much time would you like them to look at your chart, and what would make the biggest impact?
simpler pie chart: people who like pie.

Remember, the point of a chart is to create a visual illustration of data. More words means less visual impact…

5. Don’t Forget About Your Colorblind Friends

Hey, don’t forget about your colorblind friends!

  • Approximately 8% of the male population and 0.5% of the female population is colorblind or color deficient. This means if 1000 people will be viewing your chart, about 45 of them may not be able to differentiate between certain colors.
  • You can still use color in your charts… Just remember that you don’t want to use color alone to convey meaning.
  • This is something we have talked about before, so if you would like to read more information about the use of color in documents, check out my previous article on the subject.

Just remember, this chart may look perfectly clear to you:

Student Satisfaction pie chart, inaccessible colors.

… but here is how it looks to someone with the most common type of color blindness:

Student Satisfaction pie chart from before, run through a color blindness filter, and the pieces are indistinguishable.

More Notes and Disclaimers

You probably know this stuff, but just in case…

  • You have seen example charts throughout this article. All the data used to create the charts is entirely fictional.
  • Do you want to learn more about how to make charts? Please attend one of my Excel: Pivot Tables, Charts and Pictures sessions listed in myTraining.

Thoughts?

Okay, what do you think Power Users? Do you think you will be able to put some of these tips to use with your data? I would love to hear from you!

Congratulations, Power Users!

Congratulations to our newest Power Users! For the full gallery, and more information about the WSU Microsoft Office Power User Program, please visit: wichita.edu/poweruser

Tera Park

Tera Park

Mary Ann Hollander

Mary Ann Hollander

John Keckeisen

John Keckeisen

Amanda Conner

Amanda Conner