Dashboards: visualization and usability (part 3/4)

In this four-part series on dashboarding, in the first article we made clear to you when a dashboard is the solution for your information needs and what this dashboard must meet to ensure that you can get the right insights from the data. Then in the second part we showed you how to convert KPIs into an information need and a data model. In this third part, we will show you how to design visualizations and a dashboard so that it is user-friendly and answers central questions.

Choosing the best visualization for your dashboard

A well-designed dashboard draws your attention to parts where action is needed. Visualization plays a role in this that should not be underestimated. People are much better able to process and appreciate images than written text. The condition, however, is that the visualization is set up in such a way that it optimally supports the message. In our example, where a manager wants to achieve 20% sales growth over last year, the dashboard includes the number of new customers and the average order value. The visualization becomes interesting when the result has a benchmark and when the use of color draws attention.

Start with the message

It is tempting to design visualizations in the tool, but our advice is not to do so! Start with the message. What is the most important thing I want to show? Who do I want to show this to? What question does this visualization answer? As we've advised before: start with a rough sketch of your idea on paper. This will maximize your creative process. Only then take your idea into the tool you use to create the visualization. This method ensures that you keep focus on the message and that your attention is not swallowed up by all the fancy features a tool offers. So first sketch out a general picture for each KPI you want to display and then choose an appropriate type of visualization. There are an extraordinary number of options available, if you need inspiration take a look here. This seems easy and an open door, but we come across visualizations every week that leave us wondering exactly what the message is.

Using color in your visualizations

Once the basis of your visualization is in place, you can expand it with shape and color. Keep in mind that color has a meaning in dashboards, so use it only for a purpose such as distinguishing between categories or clarifying a trend, and keep color blindness in mind. Over 10% of the world's population is color blind, and choosing some colors that go well together is harder than you think. Choosing a color palette to build visualizations is best left to experts. We therefore make extensive use of a color generator that takes color blindness into account and chooses colors that match well, like the example below.

In many cases you are limited in your freedom of choice because you are dependent on the corporate identity colors of your company or client. If these corporate colors prevent you from communicating a message effectively, consult with the person responsible for the corporate identity. Often you can achieve a lot by adjusting the saturation and hue.

Formatting your visualizations

In addition to color, the formatting of your visualizations plays an important role in getting the message across. It is especially necessary that there be few distractions. Therefore, use one font, the same font size between visualizations and use accents (bold, italics) only when necessary, for example in a title. Also, do not focus on adding additional features to a visualization, but rather ask what parts you would remove more of. The fewer distractions the better, like the example below.

Source: Courtney Jordan, 'Make your data speak for itself'

Guide your user

By now you know our position on dashboards, it must:

• answer central questions;
• Draw attention to parts where attention is needed;
• lead to action.

For dashboard construction, this means that the first screen shows the status of KPIs: Where are things going well, where is attention needed? Next, a user should be invited to explore root causes. Why is the average number of orders per customer lower than last year? You do this by using mouse-over tooltips with additional information or by incorporating navigation. With the latter, users click on a visualization to see a more in-depth look at the figures. This way you provide the user with information so he or she can take action.

Use a basic layout

As a rule we work for clients with a wireframe, this is a basic layout/sketch of the dasboard where filters, navigation, explanations and legends have the same place. Again, this starts on paper and after coordination we take this design into the tool. There is no basic design that works best for every situation. Below you can see an example, however, more important is to work consistently. This makes it easier for end users to use and understand a dashboard. Place titles, buttons and filters in the same place when your dashboard consists of multiple tabs. And don't forget to explain the "hidden" features. Such as interactions between visualizations or tooltips. With a symbol or short explanation in the subtitle, it is immediately clear that a visualization has these options.

Sample dashboard wireframe

If you keep adding new filters during the construction of your wireframe or dashboard, you might wonder if the client doesn't need a dashboard but an analyst. A long battery of possible filters is confusing for many users and leads to confusion. User A sees different results than user B because a slightly different filter selection was chosen. In our experience, the limit is 5 filters in a dashboard. Do you need more filters: then you either have a complex business model or the goals of the dashboard are not as sharp as you initially thought. In our first article in the series, we discuss in detail whether a dashboard is the best solution for your question.

About data visualization and dashboarding, we personally don't run out of things to say. Want to know more? Then visit our website or sign up for our data visualization training or a hands-on training in PowerBI.

Stay tuned!
Want to be notified of the next articles on this topic? Follow us on LinkedIn or sign up for our newsletter.

Stay tuned >

Contact
Want to know more about this topic? Then contact Wouter van Gils using the contact information below.

Wouter van Gils, Senior Consultant

+31 6 15 46 99 17

w.v.gils@cmotions.nl