Usability, ease of navigation and the overall user experience are all very important in analytics dashboard design–but equally important is understanding what the end user’s objectives are when he or she uses an analytics dashboard or a user interface–because if you don’t know what they want to get out of the analytics, you risk losing the user.

Missing a target on an analytics interface can have serious repercussions. Entire analytics projects have been shelved–not because of the business value of what an app can do–but because of a poorly designed user interface or dashboard that was impossible to understand or to navigate to results.

How do you avoid this pitfall?

Sit down with your target user to determine exactly what types of answers from data they expect to get.

There is a business process behind how the user cuts into analytics results just as there is a process on the analytics side to extract meaning out of raw data.

To get to answers, the user needs instrumentation on their dashboard that enables rapid time to results. They might want to know in an eye shot how much sales are ahead of or behind target at this time of the month, or if there are any slowdowns in a city subway system. If there is problem with sales or the subway system, the dashboard or interface into the application must rapidly direct them to an actionable answer.

This sounds simple, but it isn’t. All too often, user interface and dashboard designers start the process by imagining how they would want to see data. They might not even be familiar with the thought or business processes that their end user uses.

SEE: Salesforce bolsters Sales Cloud with more Einstein AI (ZDNet)

If you’re new to interface and dashboard designs for analytics, start with the basics and use best practices.

A good entry point for beginners is to use tools like Google Analytics Help which illustrates and explains the practices and the steps for designing analytics dashboards with widgets, charts, geographical shapes and data. These best practices illustrate how to construct a dashboard or a user interface that is static and not interactive, but they give a good foundation. Common best practices include not putting too much data on dashboards so as to overwhelm the user, and selecting meaningful pictorial objects that can “tell a story” to a user in an eyeshot.

Once you understand the business process you’re supporting, move to more interactive analytics dashboards.

The big trend in analytics dashboard design in 2017 is to make them much more interactive with their users, so that users can readily drill down to get at more detailed information when they need it. Here are some showcase examples:

Nevertheless, the caveat remains that you should first understand your users’ thought and business flows. Your goal is to create an interactive dashboard that can get the user to results with minimum effort.

To illustrate, if a farmer wants to know which areas of his field are producing crops at lower yields, he wants to be able to drill down to check the soil content of lower yield areas so he can determine if a different fertilizer mix is needed. He doesn’t necessarily need to know every characteristic of his field–unless he asks for it.

SEE: DataGravity for Virtualization: Discover, analyse, protect and recover sensitive data (Tech Pro Research)

Keep it simple

One key to designing dashboards is keeping them simple and even unobtrusive. Simplicity of design is often captured by displaying only the top-level information that the user wants to know, and then giving them navigational tools that enable deeper probes if they want them. There are even cases where users don’t even want to hear from their analytics, unless there is an alert situation that triggers a dashboard.

This idea was summed up in a publication by a team of researchers who designed a dashboard for Zambian health workers to report and track incidences of malaria: “Choosing nimble tools that allow users to create and manipulate dashboards in an iterative fashion is optimal. When users have more control of the data, can experiment with the tools, and can guide development, they not only develop better products, they simultaneously build their own competency and insight.”