Author: Jeff Busch
A Brief History
Recently, I was curious about where the term dashboard came from. I know we get the business intelligence (BI) term from vehicle dashboards, but I could not remember ever learning why we call them dashboards instead of instrument panels or vehicle performance information readouts or something like that. Webster’s Dictionary to the rescue. 
Well, technically, it was Google, then Wikipedia, and then Webster’s Dictionary to make sure Wikipedia was being honest with me. Anyway, Webster’s Dictionary says that the term dashboard came from the wooden panel used to protect carriage drivers from mud splashed (or dashed) up from the hooves of the horses. Hence, a dashboard. Later, the board did double duty as something to mount gauges on for early cars. I have to admit I chuckled a bit when I read this because I realized that Business Intelligence (BI) dashboards do the same thing, or at least they should. A dashboard report will have the greatest benefit when it provides important information to the user and, most importantly, summarizes that information to hide the ugly details. These ugly details are the “mud” of the data world. Therefore, as an executive, a good dashboard still keeps the mud off your pants, so to speak. I think we can reexamine vehicle dashboards from a fresh perspective and use the analogy to gain some insight into how to design effective dashboards for our BI environments.
Dashboards – What they aren’t
Dashboards are not running boards. The running boards are what you step on to get into the vehicle. I have seen many landing pages, home pages, default reports, etc. called dashboards. However, a dashboard is not always the first screen that opens when you go into your BI website, or the page that links you to other areas of the site. A page that has several prompt controls allowing the user to apply filters to a set of common reports and then launch those reports from a link menu is not a dashboard. Dashboards can be delivered directly to users through email or mobile devices and may not have links to anywhere else.
Dashboards are not navigation (NAV) systems. The NAV system, or infotainment system, gives directions to the driver and often has features for the entertainment system, vehicle performance tuning, etc. A report with lots of flashy graphics or a high degree of interactivity is not necessarily a dashboard. Dashboards are not intended to entertain the user; they can appear very plain and simple and still be very effective. Often, simple is better. Bad things can happen when you are driving and you start poking around in your NAV system too much. Similarly, executives can easily get distracted by over-analyzing data in complex “dashboard” reports. Complex reports and detailed analysis are important ways to discover the “whys” and the “hows”, but dashboards more often focus on “what” is happening now.
Dashboards – What they are
Dashboards are BI reports that provide high level, high priority, and at-a-glance information. Often, the users are executives who do not have the time or the desire to closely examine a report to interpret what it means to them. I had a graduate school professor who said that the perfect executive dashboar
d was a page with only a green up arrow or a red down arrow. This might be a little idealistic, but the idea is that if everything is good then no more details are needed and if something is bad, a different report will be used to examine why.
A dashboard is high level and information should be summarized at the highest level possible. An executive wants to know, “Is my company on track? Is it headed in the right direction or not?” Maybe a multi-regional company would show status for each region, but getting into too much detail could obscure the overall status. If I’m driving the company like a horse-drawn carriage and I’m too interested in the details, if I’m staring at the horse’s hooves, then I could lose sight of where we are going. For example, if Product A is showing good and steady performance, Product B is doing well but trending downward, and Product C is below target but improving, what is the overall status? This analysis should be designed into the dashboard using business rules, and not imposed on the user.
A dashboard is high priority. A dashboard should show only information that answers key executive questions. Ideally, all key metrics would be combined into one “score” for the level of detail shown in the dashboard. It is important to minimize the clutter. The CEO and COO probably do not need metrics on employee headcount, and they probably do not even need company expense and sales numbers. They want to see profit, a single number, with maybe an indicator of whether it is increasing or decreasing. Vehicle dashboards do not show all information about a car. They just answer the most important questions: How fast am I going? Do I have enough fuel to get me there? Is my engine about to blow up because it is overheating? Many parts in a car spin at different rates. There is only one tachometer on the dashboard showing RPM — because that is the one the driver needs to drive properly. It might be interesting to know the RPM of the cooling fan, but it’s not really important. It’s part of the “mud”, the dirt and details of the operation that are important to someone, probably the mechanic if there is a problem, but not the driver.
Dashboards are understandable at-a-glance. One important feature of the best vehicle dashboards is that you can quickly glance at them to see the information you want. More cars are even putting key information right in front of you in heads-up displays. BI dashboards are no different. One glance is all it should take to get all the information the executive needs. Careful design choices for colors and graphics will help achieve this. Simplicity is important, as well as minimizing the use of text and numbers. One way to do this is to show a goal or threshold in a chart. Anything in the good zone is good; the specific value does not matter in this c
ase. When my gas gauge gets into the yellow area close to the bright red “E”, I know I should get gas. I am not really interested in exactly how many ounces of gas I have left. There is no such thing as only a little bit empty. I’m still walking. An interesting fact to note is that when the digital age began, numeric LED speedometers became very popular. However, people soon realized that it takes more time to interpret a number than the position of a needle on a gauge. Now even in cars that have digital dashboards with no physical dials, the speedometer, tachometer, etc. are still standard-looking gauges. Just as a driver should pull over if they need to analyze something on their dashboard more closely, like the radio or NAV system, an executive should have more detailed reports available for when they decide to slow down and look at things more closely. But the dashboard that is right in front of them is not the place for these details.
Dashboards are an important part of a complete Business Intelligence environment, just as they are an important part of a complete car. But in the same way that vehicle manufacturers have learned to keep their dashboards simple and clutter free, we need to apply the same design approach to our BI dashboards. Remember: the dashboard was invented to keep the horses’, uh, “stuff” off the drivers so they could focus on where they were going. So let’s do our companies’ “drivers” a favor and design our dashboard well enough to keep the mud off their pants.
FYI Solutions has been providing Business Analytics solutions for over 30 years. Contact us for more information regarding your strategy for business analytics and how to get the most business value from your dashboards.
