7 tips for successful dashboard design

Keep the boundaries on a single screen

People have the ability to hold onto portions of information for a short period of time, which is stored in our short-term memory. The problem occurs when an obstacle is placed in front of this information via a scroll bar or page/tab switching. Relying on the mind’s eye to remember information that is no longer visible on screen is a big ask, one that often leads in failure.

Don’t be too precise

Too much detail in dashboards or being too precise with figures will slow down the users interaction without providing any real benefit. (For example, measure charts displaying £3,767,408.78 rather than £3.7m). In design, something that does not provide a benefit is a hindrance and this especially applies to dashboard design. Every piece of unnecessary information results in time wasted by the user trying to filter out what’s important.

Choose the right display media

This is one of the most common mistakes made in both dashboard design and all forms of data presentation. Basic examples include using a bar graph when a table of data would have been more useful and vice-versa.

A more detailed example is the misuse of the pie chart.

The pie chart above represents the statistics of men suffering from cancer before they are 75. The problem with pie charts is that they are designed to represent parts of a whole, and the whole should always add up to 100%. However, with the example above, “bladder cancer” appears to take up 35% of the whole – a long way off 15.5%.

Even when a pie chart is used to represent parts of a whole correctly, the segments are not totally clear and time is still needed for the user to work out the individual statistics.

Use well designed display media

If using the wrong chart type was not enough, you must also use a well designed chart type. Most graph types used today are poorly designed as not many designers have received the correct training and lack the fundamental principles and practices for effective dashboard design.

An example is the misuse of legends. Often, legends can appear detached from the chart, usually on the right side of the screen. This is bad practise as it forces the eye to bounce back and forth, resulting in wasted effort.

This stress can easily be avoided by placing the labels directly with the slices.

Colour choice is also a common mistake in the use of dashboards. Similar colours are often used to either compliment the application theme or to prevent a high contrast, ‘child-like’ chart being displayed.

The chart above is a good example of the difficulty a user faces when trying to distinguish between ITV popular TV shows.

Effectively highlight important data

When you look at a dashboard, the most important data should immediately draw your eyes to it, even if it is not positioned in the most prominent area of the screen.

If there is no visual order to a multi-chart dashboard, nothing stands out. Everything that warrants a place on a dashboard is important, but not equally, so enhance the most important chart so the eye views the important information first.

Do not clutter the chart or its surroundings

With different design trends come greater problems. Many dashboards contain useless decoration. Bevels here, embosses there are sure to make the dashboard more appealing, however they simply distract from the data. Dashboards are not designed to win artistic awards, they are created to extract data and statistics quickly and easily and any enhancements simply prevent the user from doing so.

Use colour correctly

Before selecting the colours for your charts or chart segments, consider their meaning and interpretation when placed next to others. Some colours are hot, some are cold, some demand our attention and some are less visible and are picked up last.

If two colours on the same chart are the same, we will often try to relate them to one another.

We assume we can use red to highlight important elements yet we immediately exclude 1% of males who suffer from Protanopia colour blindness.

There are many tools available for testing and selecting the right colour, below is one of my preferred sites:

http://www.colblindor.com/color-blindness-tools/

The origin of the points above is from the book: Information Dashboard Design which is a very useful book for any person studying, developing or designing dashboards.