Dashboards are used to present large amounts of information in a condensed and visual form.
They are meant to be represented on a single screen, to make an efficient use of the space, avoid superfluous graphics and include actionable, meaningful and relevant data according to viewers objectives.
However, we can find some examples where dashboards display the information in the wrong way. These are some examples of bad designs that can make a nicely intentioned dashboards fail.
1. Tiny and spaced
Yet the dashboard is displayed in one single screen, the amount of white space between charts make it difficult to read as unique piece of information. Increasing font-size and adjusting proportions would make it more efficient.
2. 3D + raw data: what for?
3D charts might look fancy, but they make any dashboard
3. Too many colours, no data visualisation
They call this thing a ‘Patient Dashboard’, a.k.a. Electronic Patient Record, and yet the goals are well aligned to what a Dashboard means, the whole design is wrong.
There’s no data visualization but tables with raw data which makes impossible to know which are the key indicators to drive decisions. Colours are totally useless and they badly contrast with the ones that are used to highlight the important data e.g. does s/he
There are some obvious fails here: poor design, lack of clear actionable data, too much information inside and outside the dashboard.
4. Lack of consistency
Consistency is such an important characteristic of a good design that when it’s broken it may cause real mistakes on people’s choices. When it comes about dashboard design, we can also find some examples
In this example we can see how colour codes are representing a variable common in different charts (Green for Abandoned, Blue for Completed), using the same colour will help to identify that variable in a new chart.
This is used consistently in the first row, however, in the second row the blue and green colours are reused to represent a different variable which creates confusion.
5. No highlighting thing (or too many)
Hedwig Von Restorff was a student at the psychological institute of the University of Berlin with Gestalt psychologist Wolfgang Kohler who made some interesting experiments where discover that people tend to remember easily things that differ from the rest by distinctive features like shape, colour, or orientation.
This discovery builds the rationale under the usability principle ‘Recognition rather than recall‘. If we design dashboards in a way that nothing stands out, people will require a bigger effort by scanning items one-by-one. Another mistake is failing by displaying important items on paces that people are used to ignoring, this is known as ‘banner blindness‘.
In this example we can see three mistakes related with this phenomenom:
- Numbers are in bold using a big font-size which indicates they’re KPIs. This style helps to identify a KPI wherever it is.
- There are such many numbers using the same style that they all seem to be equally important.
- The text between the first row of KPIs and the rest of the charts it’s hard to read and it makes difficult to identify the relevance and meaning of the content compared with the rest of the dashboard.
6. Fragmenting the analytical experience
Scrolling is a way of fragmenting the information displayed on a dashboard, and fragmenting information interrupts the experience of analysing data. This interruption may cause frustration and also the abandoned of the analytical activity which is counterproductive.
The consequence of this bad design choice are multiple:
- Relevant information could be hidden. Instead of using a scrollbar inside a ‘widget’, the information could be organised in a way that let users to drill-down to get more details.
- Scrollbars are difficult to handle when they’re close to each other.
- If the dashboard doesn’t clearly invite to interact, scrollbars can be ignored.
Dashboards should facilitate comparative analysis and gaining insights into the represented data. In order to design a valuable
This article summarised the exercises I made during the course: Information Visualization: Getting Dashboards Right.