9 Tips for Building Actionable Data Visualizations

By March 6, 2017Tips and Tricks

The power of data comes from the analysis itself. However, the power of the analysis depends on how that data is presented. An analyst can pour hundreds of hours into the depths of Excel madness, but ultimately all that matters are the insights that are delivered to the business leaders. Good data visualizations have the ability to convey key insights more effectively, while bad visualizations have the ability to derail a presentation. Or worse – inaction on critical learnings.

We often see artsy and beautiful visualizations but they are not necessarily speaking to decision makers. Visualizations such as the one below definitely have a wow factor but can we discern an actionable insight? To create practical data visualizations for business, the most important goal is to efficiently present information in a way that demonstrates the key takeaways clearly. This type of [practical] data visualization is the intersection of data, design, and psychology.

Beautiful chart, but is it actionable?

Beautiful chart, but is it actionable?

Below are a few MaassMedia best practices on how to create practical data visualizations. The information is not only a compilation based on design and psychology research, but also from our own experience presenting to dozens of clients across verticals.  A few minor visualization adjustments can significantly clarify your message. Additionally, for some of our examples below, we will also be utilizing characters one of our all-time favorite workplace TV shows, The Office.

 

1. Remove Noise

It is important to avoid drawing attention to irrelevant elements such as gridlines, background, and excessive labeling. The goal is to display a high “data-ink ratio:”

(Data ink) / (Total ink to print the graph)

Elements that reduce whitespace without adding analytic value would fall into the category of “total ink to print the graph.” Removing noise also helps create figure/ground contrast so that it is easier for the audience to focus on the data points of the graph rather than on the irrelevant background.

Before:

After:

Tells the same information as above but much cleaner and easier to digest.

 

2. Use Round Numbers

Long numbers are hard to comprehend. Our mind loves things that are simple and clear as it takes less time for us to process the information. While it is important to provide detail, adding decimal places will not change the story.

Before:

Insight: “Dunder Mifflin sold 14,184,419 paper reams in Q4.”

After:

Insight: “Dunder Mifflin sold 14M paper reams in Q4.”

 

3. Pay Attention to Positioning

Ensure that labels are easy to read and not misleading. Labels should be horizontal – and if there isn’t enough space for the labels on a vertical bar graph, switch it to a horizontal bar graph. Sometimes it is difficult to properly gauge axes so labels should be added at the end of the bar. For instance, if the bar ends in between the major marks, estimation is required. Additionally, sort and order data appropriately, either alphabetically or sequentially by value as our brains process simple and orderly patterns faster.

Before:

Insight: “Creed had the most pairs of cool shoes.” Reading the labels requires the audience to tilt their heads. Also, in order to figure out who has the most cool shoes, the audience has to scan past 4 other names.

Working:

Insight: “Creed had the most pairs of cool shoes.” Now that we know Creed has the most shoes, what’s the difference in the number of shoes that Creed and Kelly has? Hard to tell…requires estimation.

After:

Insight: “Creed had the most pairs of cool shoes.” This is easy for the audience to figure out that Creed has 4 pairs of cool shoes more than Kelly.

 

4. Use Proximity for Comparisons

Objects placed close to one another are perceived as a group. Similarly, objects further apart seem less related. If your most critical insight isn’t immediately clear, re-group the data to prevent it from competing with tertiary conclusions.

Before:

Insight: “Dwight’s Dundie collection was the only one that grew quarter over quarter.” Difficult to tell from this graph… the audience would need to look at each employees’ Dundies and compare across the separate [quarter] groups.

Insight: “Dwight’s Dundie collection was the only one that grew quarter over quarter.” Difficult to tell from this graph… the audience would need to look at each employees’ Dundies and compare across the separate [quarter] groups.

After:

Insight: “Dwight’s Dundie collection was the only one that grew quarter over quarter.” Significantly easier to determine the trend of Dundies received by each quarter over quarter.

Insight: “Dwight’s Dundie collection was the only one that grew quarter over quarter.” Significantly easier to determine the trend of Dundies received by each quarter over quarter.

 

5. Add Emphasis with Shapes

Shapes help guide the audience’s eye immediately. Use shapes to create a distinct focal point.

Insight: “Dwight’s Dundie collection was the only one that grew quarter over quarter.” Where’s Dwight on this graph? It takes the audience extra time to identify.

Insight: “Dwight’s Dundie collection was the only one that grew quarter over quarter.” Where’s Dwight on this graph? It takes the audience extra time to identify.

After:

Insight: “Dwight’s Dundie collection was the only one that grew quarter over quarter.” Call out Dwight with a shape.

Insight: “Dwight’s Dundie collection was the only one that grew quarter over quarter.” Call out Dwight with a shape.

 

An alternative would be to call out Dwight with a background shape and utilize an arrow to highlight the QoQ growth.

An alternative would be to call out Dwight with a background shape and utilize an arrow to highlight the QoQ growth.

 

6. Create Focus with Color

Use color sparingly and consistently. An accent color is great for highlighting a significant data point. However, be mindful of red, green and colors with certain connotations. For instance, if a brand is associated with one color, use that color consistently throughout the presentation to represent that brand. Various shades of grey are recommended if additional differentiation in the graphs are needed.

Before:

Insight: “Even though Stanley always wears the same sweater, he actually owns 8 different ones.” Where’s Stanley on the graph? It takes the audience extra time to identify.

Insight: “Even though Stanley always wears the same sweater, he actually owns 8 different ones.” Where’s Stanley on the graph? It takes the audience extra time to identify.

After:

Insight: “Even though Stanley always wears the same sweater, he actually owns 8 different ones.” Call Stanley out with a pop of color.

Insight: “Even though Stanley always wears the same sweater, he actually owns 8 different ones.” Call Stanley out with a pop of color.

 

7. Be Cautious of Pie Charts

The audience can only accurately gauge the size of pie slices if they are in familiar percentages (25%, 50%, 75%, 100%). Thus, it is difficult to compare other sizes effectively. Similar to pie charts, graphs with other special effects like 3D distorts the viewers’ ability to analyze size and length properly.

Before:

Insight: “Darryl consumed the most apples.” Did Darryl actually consume the most apples? From the graph, it looks like Michael consumed the same amount.

Insight: “Darryl consumed the most apples.” Did Darryl actually consume the most apples? From the graph, it looks like Michael consumed the same amount.

After:

Insight: “Darryl consumed the most apples.” With a bar chart, it is clear that Darryl did consume more apples than Michael.

Insight: “Darryl consumed the most apples.” With a bar chart, it is clear that Darryl did consume more apples than Michael.

 

8. Limit the Categories Displayed

Exclude unnecessary categories. For pie charts, visualize no more than 5 categories per chart. Utilize an “other” category if needed. For line charts, plot no more than 4 lines per chart. Break up the chart into separate graphs for better comparison. If a category does not add to the story, always ask why it’s critical to the story.

Before:

Insight: “Wait… what is even the insight here?” Too many categories, no trends seem to stick out.

Insight: “Wait… what is even the insight here?” Too many categories, no trends seem to stick out.

After:

Insight: “The consumption of healthy snacks decreased over the weeks." Separating the line graph into categories allowed us to tease out an insight on healthy snacks.

Insight: “The consumption of healthy snacks decreased over the weeks.” Separating the line graph into categories allowed us to tease out an insight on healthy snacks.

 

Insight: “Consumption of chocolate bars is correlated with that of candy packs.” Similarly, another insight was able to be teased out.

Insight: “Consumption of chocolate bars is correlated with that of candy packs.” Similarly, another insight was able to be teased out.

 

9. Choose the right graph

Using the wrong graph can make it harder for the audience to digest the information. Below is a handy guide from Andrew Abela.

choose-the-right-graph_abela

 

Now that you have the basics of practical data visualization under your belt, go put them into practice! While it is rare that the audience will praise the effectiveness of your data visualization, bad visualizations will certainly garner comments of confusion. Practical data visualizations should not be the focus of your presentation, but are there to seamlessly and naturally support the key insights. In conclusion, the tips discussed above all fall largely into two main concepts – simple is always better and call out key insights.

Want to receive the latest in our continuing series about improving data visualization and analytics insights? Subscribe to MaassMedia Insights.

One Comment

  • JAMSO says:

    Great article and overview that will help many deliver improved impacts.
    These are important parts of the full story. The full story is how, when and to whom the presentations and information is shared.
    Clear follow up actions need to be encourage otherwise your charts simply inform and create lower potential value.
    Some extra tips to address that are as follows:
    1) Number your slides and create minutes with action next to each slide
    2) Choice of audience is important and seek methods of engagement where possible and if needed.
    3) Create an positive introduction, short overview and take away for each slide. Do you wish to shock, stimulate or inform?
    4) Practice, do not expect to get it right overnight. Significant rapid improvements can be made from the content shared within this article but story line and narrative might take longer to master.
    5) Take a peek at some tips for public speaking !

Leave a Reply