by Mike Hanson
Date Published November 17, 2015 - Last Updated 7 Years, 290 Days, 11 Hours, 53 Minutes ago

If you’re in a tech support leadership role, odds are good that you have been asked to create a dashboard. The person making this request might not have a clear idea of what they want to see. They just know that when something needs to be monitored or tracked, a dashboard is a common way to meet that requirement. If you have data, somebody will probably ask for a dashboard. As leaders, we need to understand how to present critical information effectively in a clear and easy-to-read format.

If you have data, somebody will probably ask for a dashboard.
Tweet: If you have data, somebody will probably ask for a dashboard. @ThinkHDI

I’m not going to focus on the data or what specific metrics you should measure. Rather, I will focus on how people perceive information. When presenting critical data, the design of the dashboard or scorecard is important. We need to understand what people see when they look at a collection of data, charts, or graphs.

We need to understand what people see when they look at a collection of data, charts, or graphs.

Dashboards have become increasingly important because modern business is driven by data. In years past, there may have only been a single data source, so accessing that information could be done quickly and easily. Today, we live in an age of big data where there are many back-end, inter-related systems, high performance databases with hundreds or thousands of data points, large numbers of queries or reports to manage, and systems distributed across the entire globe. Accessing these systems and data is far from simple, and understanding the relationships and tying data together in one place is what has driven the increased interest in dashboards.

A dashboard is usually—though not always—a visual mechanism for displaying business-critical information. Predecessors of our modern dashboards would be things like executive summaries, Executive Information Systems (EIS), or Online Analytical Processes (OLAP). Today, dashboards provide ready access to vital KPIs, goals, and other important metrics. Dashboards typically collate and consolidate data from multiple, disparate systems into a quick, at-a-glance visual representation. Dashboards often come integrated with many of the tools we use in IT support, and, of course, dedicated products can be used to build them.

Communication Is Key

First and foremost, a dashboard is a form of communication. In most cases it should be a single page. It should be easy to read and interpret, offering clear summaries of the data and establishing a clear and concise context for the presentation.

There are really three different types of dashboards, each having a different purpose and presenting data in a different way. The first is the strategic dashboard. This type does not need to contain real-time data, as it is used to examine long-term trends and enable decision-making. It primarily focuses on high-level, historical data that can in turn be summarized for leadership, so they can clearly see where the organization has been and where it is going.

Next is the operational dashboard. This is the type that most mid-level managers are familiar with, as it is tactical in nature and offers real- or near-real-time views into the data. It shows how a team is doing right now, and enables management to adapt to and manage dynamic business environments. Because operational environments are typically very busy and high-stress, these types of dashboards need to be easy-to-read summaries, with highlights on the status of critical business functions. Managers need to be aware when things in the environment change so they can provide an immediate response.

Last is the analytical dashboard. This type of display is more complex than the strategic or operational control panel. Its purpose is not to drive decisions, but rather to drive investigation. As a result, it tends to be less graphical in nature and more data-driven. Containing more numeric or textual components than a typical dashboard, it provides views into lots of data, showing the detail and complex relationships not used by other types. This dashboard would not be read at-a-glance, but rather would be closely examined and analyzed in detail.

Avoid Common Mistakes

When putting together a dashboard, avoid some common mistakes. Just because data can be summarized, doesn’t necessarily mean that it is useful or critical to running the business. Avoid dashboards just for the sake of having one. To use visual or graphical displays, we need to understand how people perceive data and then design a dashboard that takes advantage of those perceptions.

The first common mistake is improperly highlighting the important data points in the dashboard. If there are numbers, is it easy to differentiate the critical data from the rest of the display. If you include a chart or graph, do the colors and text call out exactly what needs to be communicated?

The next mistake, ignoring context, is closely related. To get useful information from a dashboard, the viewer must not only be able to quickly see the relevant data, they need to be able to understand what they are seeing and why it is important. If there is a critical statistic in a long list of similar numbers, make sure that it’s not only highlighted but that the reader can understand why it is critical. Is the context, or bigger picture, clearly evident?

One of the hardest things to overcome is not a design issue but really human nature. Everyone has a pre-conceived bias, and as a result, people tend to see what they expect or want to see. The Disney theme parks provide a great, real-world example of this notion. Images of Mickey Mouse that blend into the surrounding landscape and architecture are hidden throughout the park. If you’re not specifically looking for them, they can be almost impossible to see. Dozens of books have been written to help visitors locate these hidden Mickeys. Similarly, important data can hide among what many people would see as a normal pattern. When we build a dashboard, we need to make sure that it effectively calls out breaks or changes in what is perceived to be the typical pattern.

When designing a dashboard, we want to keep in mind the primary goal of keeping it clear and easy to read. Don’t put dark text on a dark background or light text on a white background. Be consistent in units of measure; don’t use percentages in one spot and decimals in another. If you have percentages or variances noted, be sure to explain it so the reader doesn’t get confused or have to spend time figuring out what the designer intended.

A dashboard doesn’t have to look like the dashboard in an automobile. In fact, extreme design features are counter-productive, so make sure that the design doesn’t overwhelm the content. Pay attention to the layout, avoid outlandish color schemes, and restrict graphics to those that actually contribute to the content of the dashboard. Remember, you’re not putting on a show; you’re trying to provide critical information for the viewer.

On strategic and operational dashboards, avoid going into too much detail. Keep it high-level with quick overviews and summaries. Avoid complex charts; for example, pie charts with too many slices can make it difficult to call out specific data points. Too much information will slow your reader down and has the potential to be very confusing. A dashboard with an overwhelming amount of data or confusing charts is self-defeating in the long run and won’t be used consistently.

Follow Basic Principles

At this point, we have an idea of what we shouldn’t do; let’s look at some simple, basic principles that will help us build useful dashboards.

First, give thoughtful consideration for the content of your dashboard. Does it address specific objectives? It should be customized for the intended audience, summarizing and isolating your objectives from any large sets of data. Instead of tables of numbers, use sums, averages, distributions, correlations, and so forth. The dashboard should consist of short, direct communication.

Consider as well something called the data-ink ratio. Once you’ve built what you think is a good dashboard, step back and examine it with a critical eye for the ratio of valid, core information—this is what you’re trying to communicate—versus the overall white space and other non-value-added content like grid lines, 3D graphs, decorations, gradients, or anything else that does not directly influence the message the dashboard is trying to communicate.

Because a goal with strategic and operational dashboards is being able to present information quickly and in a single glance, it is important to confine the information you’re trying to share to a single screen. Avoid fragmentation, which requires the reader to navigate to a different screen for additional information; likewise, make sure all of the data appears above the fold, formatted in such a way that none of the information is out of site and requires scrolling on the part of the reader. Having this data cut off sends a message that it is not important. Note that analytical dashboards are an exception to this rule, as they typically need to go into more detail, and often offer the ability to drill down into supporting data.

Keep the context of your dashboard in mind as well. As noted above, this is a common failure point, so keep in mind that data without context can be misleading or even completely meaningless. For example, is there a graph displayed on the dashboard? Be sure that the x and y axis are clearly and consistently labelled. Any numbers need to clearly communicate their meaning, whether it’s good or bad.

Give consideration to your formatting. Closely examine what needs to be communicated and determine if the message would be better as a set of numbers or some sort of chart or graph. When you use graphs, try different types to find out if one kind works better than others. It’s very important to never assume your reader will understand or comprehend data points or charts simply because they seem to be clear to you. Label everything, and add descriptions whenever possible.

How the dashboard is organized can help your viewers understand what is most important. The data should be arranged in the way it should be used, with important and critical information prominent on the page. Highlight items using larger fonts, different colors, or bold text. If the point of a data set is to compare it with other data, is it easy to do that? Will your reader be able to understand what needs to be compared and why?

Since the most common types of dashboards are strategic and operational, most of the suggestions above will tend to work better with those types. Analytical dashboards are less common, and often more detail is required. Unlike the others, an analytical dashboard will often have the capability of drilling down into further detail. The front page of an analytical dashboard should contain summarized views and any relevant graphics; from there, the user can follow links or click on data to move down into a multidimensional view that shows the data that has been collected and its source. If necessary, you can drill down further to display detailed reports, specific transactions, or operational information.

An internet search can be very helpful to see examples of great dashboards. Also, consider the dashboard in an automobile. There’s a reason why the old-style analog gauges have lasted so long. Digital displays have been tried many times over the years, but nothing gives a quick, at-a-glance picture of how the vehicle is performing better than a set of needles. Even recent LED displays have abandoned sets of numbers to emulate the old-style gauges or to simply provide additional information for the driver.

Evaluate Tools to Build Dashboards

Plenty of tools are available for building an effective dashboard. They range from a desktop spreadsheet to custom-built solutions. Each approach has pros and cons.

Using a program like Microsoft Excel to build a dashboard is a good choice because it’s probably a tool that’s readily available. It has a good graphing engine, and some nice plug-ins are available to expand its capabilities. The down side is that it’s not easily portable to the web, and its data-handling capability is somewhat limited.

There are dedicated reporting tools like Crystal Reports or Tableau that offer dependable and customizable options. Often they have canned reports or templates that can assist in building a dashboard. Unfortunately, these tools may not be an option for smaller teams or businesses due to costs or the lack of staff familiar with the programming skills needed to use them.

If you use a service desk tool, you likely have built-in analytics or additional modules that can be added to give the functionality to create a dashboard. Like Excel, it’s probably something already in the toolbox, and because it’s a normal part of the support process, it should be easy to integrate into the overall workflow. However, this solution also probably requires some development knowledge, and the types of data displayed may be limited to what’s provided within the service desk application, limiting access to outside data.

Finally, companies like PureShare or Xtraction offer dedicated systems for building dashboards . These can offer some fantastic solutions, because they are highly focused on effectively displaying data. They often have a dedicated server and can jumpstart a project by offering some great prebuilt dashboards or offer expertise on creating custom views. Like some of the others, these tools require access to staff with development knowledge, and often the costs can be prohibitive.

Focus on the Purpose

The need for dashboards will only increase as our access to data continues to grow. Our ability to gather and present this data in a way that our customers can clearly understand is vitally important. Hopefully this article has provided some guidance in doing that.


Mike Hanson has been involved with many aspects of IT over the past twenty-five years, from application development to desktop support. Today, he is a technology support director at Optum, Inc., where he proactively seeks ways to improve service delivery for more than 100,000 clients, through process improvement, knowledge and problem management, demand management, asset management, metrics, communication, and training. He is a current member and past chair of the HDI Desktop Support Advisory Board, VP of Content for the HDI Skyway local chapter, and a certified HDI Support Center Manager. He holds ITIL Foundations and Practitioner certifications.

Tag(s): balanced scorecard, business intelligence, dashboards, KPI, metrics and measurements, performance management, Reporting and Analytics, reporting-and-analytics, reporting, service management, support center, supportworld, tools


More from Mike Hanson :