Top 10 pitfalls in Dashboard Design

Dashboards are one of the most effective ways to make information available to the business so that performance can be monitored. Using a great Business Intelligence (BI) tool does not automatically result in effective dashboards. The design of such a dashboard must take into account the end users, the information they want to see and how that information can be displayed most clearly. However, this process is prone to a number of pitfalls that you need to avoid. Below are the top 10 pitfalls in dashboard design, or the “Common Pitfalls” described by Stephen Few, author of ‘Information Dashboard Design’

#1: Exceeding the boundaries of a single screen

One mistake made in dashboard design is making the user scroll. It is much more powerful to place information together. That makes the information easier for the user to process and saves time. By placing information on one page, or sheet, the user gets a complete overview. This allows them to discover important connections in the data that might otherwise not have been so clear.

#2: No context for the data

Context can make the difference between meaningless numbers on a screen and numbers that give people insight and that they can act on. Without context, the user does not know whether a measure shows an improvement or a deterioration of the situation. Measures are enriched by context in the form of, for example, a comparison, target or history. A good example of a lack of context in a dashboard is a ‘gauge meter’; without a comparison or target, a user does not know whether this visualization indicates a deterioration or an improvement.

#3: Showing too much detail or precision

To keep the information in a dashboard easy to take in, there is no need to show too much detail or precision. When you do, the user is forced to process the data at a detailed level, even though this is often not relevant to their purpose. Think, for example, of the number of decimal places. There is no need to show down to the second when a task was created.

#4: Choosing the right measure

Measures must be aligned with the needs of users and how they want to use this information. Users must be able to see clearly what is being measured and in which unit it is expressed. If this is not taken into account, information can be difficult to process, as you can see in the example at the bottom left.

#5: Choosing the right chart

Choosing the right visualization is where most people go wrong. When designing a dashboard, you should first look at the data. Which chart best represents the information? People often go wrong here by, for example, choosing a pie chart, while differences in quantities are often clearer in a bar chart.

#6: Unnecessary variety

When a dashboard uses an unnecessary variety of charts, you lose the overview. Developers sometimes try to avoid having the same chart type appear several times in a dashboard (for example a bar chart) by deliberately using other charts. The danger here is that the chosen chart may not represent the data best.

#7: Inaccurate quantitative data

When quantitative data is displayed inaccurately, it can give a distorted picture of the situation. In the example below, the Y axis starts at $500,000, making it look as if revenue is 4 times as high as costs in January. Because the Y axis does not start at zero, it becomes difficult to see that revenue is actually only twice as high.

#8: Arranging information poorly

When you design a dashboard, you cannot just place charts in any random order. Two things are essential here: 1) important information in a dashboard should stand out more than less important information, and 2) information that should be compared must be placed together.

#9: Avoid useless decoration

Decoration in a dashboard that serves no purpose is useless and only distracts from the important information. Leaving white space in a dashboard makes the message clear more quickly and gives the dashboard a calmer look. This allows the user to process information faster.

#10: Use of color

Color can be used to show differences, emphasize certain information or show the relationship between pieces of information. Colors must be chosen carefully based on how people perceive color and the significance of color differences, such as warm and cool colors. It is also important to take color blindness into account; around 10% of men and 1% of women are affected by it.

Stephen Few explains that Business Intelligence can only add value to the business when a company's most valuable asset, human intelligence, is engaged and stimulated by it. Communicating data visually is only effective when the design of dashboards takes into account the way people see and think. It is important that a dashboard contains all the necessary information in the most condensed way, without losing anything. It must also be presented well visually so that users can easily read and understand it. When these points are taken into account in dashboard design, the information shown becomes easier for users to process. As a result, they can quickly gain important insights and take the corresponding actions.