Posted By Dashboard Spy / 14th September 2011
A Dashboards.org reader sent me this article to post: Doing Dashboards Right
Software dashboards that consolidate and report performance metrics about programs, business processes or information technology infrastructure can be valuable tools for agency managers and decision-makers. Dashboards can also become seldom-used frills if executives decide they don’t deliver the information they need to do their jobs.
Experts say that a dashboard’s success or failure boils down to a few key elements.
The right metrics
“Before you even go buy the technology, know what you are trying to accomplish and examine if you’re going to be able to do it,” said Greg Cohen, chief of business management and metrics at the Coast Guard’s Acquisition Directorate. “And then you have to stick with it and take action on it, or people won’t take their metrics seriously.”
The right tool
It can help immensely if the dashboard software can incorporate many different data types. Technologies that do not require a lot of up-front schema knowledge or data modeling tend to work the best, said Matt Eichner, vice president of strategic development at Endeca.
Older dashboard systems required data modeling that had to anticipate what questions the dashboard users might ask. New systems are capable of finding data relationships automatically.
“In the intelligence and defense communities, the ability to do off-road analysis is a fundamental requirement,” Eichner said. “So dashboards must provide exploratory capabilities that any person can use. They must provide high-level summaries but also guide users through the next step process of drilling into the individual pieces behind it.”
Agencies might also want a dashboard that can collect data automatically. Manual collection increases costs dramatically and can become a major burden. Because dashboards are only as good as the data going into them, determining how the data will be collected is critical.
The right processes
Starting with a small trial project is a good way for agencies to decide whether dashboards are a good option for them and work out the kinks before a bigger implementation, said Shawn O’Rourke, vice president of risk management services for American Systems, a government technology solutions provider.
“Do not try to capture all your information on the dashboard the first time out,” O’Rourke said.
Also, don’t underestimate the potential resistance to a program that will implicitly monitor people’s performance, O’Rourke said.
“When instituting a dashboard, it is important to engage the organization to communicate the dashboard’s purpose within the organization,” he said. “Open communication and clear goals will help ensure the trust in the program and its operations. Also, test driving the dashboard before it goes into full operation will help bring attention to any potential problems.”
Recommended Reading
Posted By Dashboard Spy / 21st June 2011
The last post on selecting colors for our business intelligence dashboard design brought out a few queries about color selection and other aspects of dashboard design.
One issue that came up was about the use of design elements to call attention to different parts of a dashboard.
Let me cross post here the following Dashboard Spy article on the use of Preattentive Variables on dashboard design:
Executive Dashboard project participants are fortunate when they have the opportunity to work with true graphic designers on the visual design portion of the dashboard. There is a real opportunity to see that design is a real discipline and that the value the visual design team is not merely skin deep, but integral to the success of the dashboard interface. Of course, graphic designers vary in skill and understanding, but you know you are in the presense of a real master when they teach you the fundamentals of their trade. When they expose the “why” of their decision making, you learn that making your business intelligence dashboard intuitive relies on understanding factors of human perception.
Let me illustrate the point with a great example from the incredible book, Designing Interfaces by Jenifer Tidwell.
In the chapter on “The Basics of Information Graphics”, the author points out a topic of great interest and value: Pre-Attentive Variables. Basically, the idea is that you have to choose your visual features very carefully because they operate “preattentively” – that is, they convey information before the user pays conscious attention.
The author walks us through a great demonstration that we replicate for you Dashboard Spy readers. Here is a series of graphics to study to understand the principle so that you can apply it to your business dashboard designs:
Look at this first figure and find the blue items.
Assuming you’re not color-blind (and probably even if you are), you can easily pick out the blue items.
Now try it with a larger field of data items:
Still not a problem, as you can see. No matter the number of balls in total, you can easily pick out the blue ones, right? In fact, if you think about it, the time it took for the larger data set was not more than that for the first smaller set. You just look at it and see the blue ones. Instead of being a linear relationship with the number of objects, the time is not dependent on the number. That’s because this example operates on color. That’s the key, because color recognition happens at a primitative cognitive level. Your visual system does the heavy lifting for you and it works in a massively parallel manner. I bet we could show you a page full of balls and it would not take any more time to notice the blue ones.
Now, for the exciting part. Take a look at a text table.
Search through the above table for values greater than one. Not so easy, is it? But the real upshot is that if I were to increase the table size, it would take you longer, wouldn’t it? In this case, our parsing of the data values IS linear with the number of values displayed.
Now take a look at this version of the data table:
Read Full Post »
Recommended Reading
Posted By Dashboard Spy / 12th March 2011
It used to be that dashboards were just plain utilitarian and ugly. Almost as if having a cheesy steering wheel metaphor taken literally was mandatory. Then development teams started using graphic designers and dashboard design started really taking off. Now we finally have some decent business intelligence applications that are thought out from the user experience point of view.
If your in-house designer or user interface expert isn’t quite as up to speed as he or she should be in terms of user experience as a separate (and rigorous!) discipline, please send them to this resource:
Human Factors International Webcasts and Podcasts
They even have full transcripts of the excellent interviews with leading user experience experts.
Some of the excellent topics offered include:
- The Usability of Electronic Health Records at Mayo Clinic
- The ROI of Usability and Making Usability Routine
- More than just a pretty page… The importance of user interface design patterns
- How to Find Your Executive Usability Champion
- The Science of Persuasive Design: Convincing is Converting
- “Recession-proof” your business through customer-centered design: their experience IS your business
- Optimizing e-commerce UX: Create websites that empower customers, are effortless to use, and create a joyful shopping experience
- Designing effective “self-serve” portals for user convenience and business efficiency: SAP – a case study
- Managing the knowledge behind business decisions through user-centered design – a case study
Recommended Reading
