Designing Dashboard : 5 Best practices
In the age of big data, machine learning and AI, the need for human friendly information visualization has increased more than ever before. Having worked on designing dashboards for a few years now and influenced by experts in the field of information visualization and user experience like Stephen few, Jakob Nielsen, etc. writing this to share what I think are the best design practices to build sleek and user-friendly dashboards.
Stephen Few defined a dashboard as:
“…a visual display of the most important information needed to achieve one or more objectives; consolidated and arranged on a single screen so the information can be monitored at a glance”.
Whilst the information visualization techniques and approaches are constrained by the raw data and the technology, the intended users should always be paramount in designers’ mind when deciding on the UI design and how to represent the raw data without sacrificing meaningfulness. Failing to produce user-friendly information visualizations on the dashboards can impede the analyst’s ability to extract the meaning underlying the text/numbers and like most dashboards, they will continue to be just the eye candy!
A major challenge for designers is to represent large data in a meaningful, coherent way which is user-friendly in interaction. The information should give comprehensive insights and it should be actionable. Below are the five things, among many other, I consider as important to design a efficient and meaningful dashboard;
If possible, avoid fragmentation of information. Fragmentation of information could happen if there are tabs or scrolling in the dashboard. Fragmentation forces the user to switch between sets of data, interrupting the analytical experience. Also, dividing the data will introduce additional steps into the process of analyzing information which could be a frustrating user experience.
Scent of Information
Usually, when the user first arrives they have a central aim, whether they want to see company performance, oversee the employee work and progress or to analyze some trends. This main aim will prime them to find the necessary related information and ignore most of the remaining information.
Chi, Pirolli, Chen, Pitkow (2003) termed this human tendency — to detect only information relevant to their current task — “following the scent of information”.
To allow this, the dashboard must be arranged in a logical and structured manner. Cluttered displays force the user to ‘search and find’ process, which is time-consuming and effortful.
Our ability to better remember or extract items that appear to stand out from the rest with greater immediacy is referred to as the Von Restorff Effect.
Highlighting is an extremely useful technique in directing users’ attention, especially when the display has a clear order of importance. Highlighting of information could be done by the use of colors, altering size and weight of text, or/and by positioning the UI elements distinctly. However, highlighting must be used sparingly; otherwise this technique loses its power to draw attention.
Consistency and Standards
Designers may feel using the same graphics for different sources of information in the same display threatens the overall aesthetic appeal, but focusing on just the aesthetics and not on the viewer’s understanding could hamper the user experience. Consistency means the user only has to learn one set of behaviors for each task, and apply them on each separate occasion. Use the same styles, labels, functions, icons and display methods to reduce the amount of new learning the user must undertake. This will save time, effort and potentially improve the user experience.
Based on research, it is well known that the user scans the screen from top left to right and top to down fashion. The most important information should be located in the top left or center of the screen and the remaining space must be ordered so as to lead the user to more specific information. The top right-hand corner and the bottom left corner should be used for progressively more detailed information. Conforming to established design structures can promote rapid detection of target information.