Dashboard development and data visualization tools for effective BI
A comprehensive collection of articles, videos and more, hand-picked by our editors
The all-new Dodge Dart features a dashboard with a stylish new lighting system and a highly customizable display. The problem is that the ambient lighting is hard to see in a brightly lit showroom, so all the "sizzle" of the new dashboard design is lost on potential car buyers. In addition, some auto buffs are already complaining that the display is too customizable and distracting.
For IT and business intelligence developers crafting BI dashboards to convey corporate data to business executives and workers, there can be similar tensions between the imperative to deliver what users need and the temptation to offer glitzy graphics, animations and effects that get in the way of effective data analysis.
David Stodder, director of BI research at The Data Warehousing Institute in Renton, Wash., said that if the intended end users are experienced with spreadsheets and sophisticated analyses, dashboard designers might want to include more complex charts, graphs and scatter plots. But more often, he added, it makes sense to start simple on BI dashboard design.
The end result, according to Stodder, should be a design that presents data clearly and tells people something useful for their specific jobs. "Studies show that [business] professionals spend a lot of their time simply trying to find the right information," he said. "The primary role of the dashboard is to cut down on that and help make employees more productive."
More in this series on dashboard design
Avoid common pitfalls in designing BI dashboards, from clutter to not fine-tuning designs for different users
Explore approaches to designing mobile dashboards that balance usability and functionality
Diving deep into dashboard data
Stodder sees interactivity as another important element of good dashboard design. "Many dashboards have been developed that are just visual versions of the canned reports that companies produce," he said. "They don't offer any drill-down capabilities to help people understand data and its ramifications." Providing such capabilities can be the most challenging aspect of building dashboards, partly because it's dependent on the functionality in underlying BI systems. But a lack of interactivity can reduce the adoption of dashboards and encourage users to return to the comfort factor of their tried-and-true spreadsheets, Stodder said.
Craig Schiff, president and CEO of consultancy BPM Partners in Stamford, Conn., said designers often focus too much on data presentation issues and miss the importance of delivering the right data to users. "They worry about color and 'stoplights' and gauges, which can all be helpful -- but they don't matter if they have the wrong information," Schiff said.
That sentiment is shared by Charlie Claxton, vice president of creative strategy at Produxs Inc., a developer of custom Web and mobile applications in Seattle. "What tends to be lost in dashboard design is the user," he said. "It amazes me how often we come across dashboards that are really just a set of disparate widgets cobbled together with little to no overarching meaning, other than they were parts of reports found in the system." Well-designed dashboards go beyond the concept of telling users "here is a lot of data" and enable them to do their jobs more successfully, Claxton added.
That means BI dashboard design teams should do as much research as possible to fully understand the roles and needs of the workers who will be using the applications. "They are your audience," he said. "You have to drill down and be clear as to the particular job function, business process and organizational thinking and how to meet that dynamic in a way that is comprehensible and of course smart."
It's good to have options on dashboard design
Getting executives and other users involved up front is a critical part of that process, said Bill Brydges, a managing director at consulting company MorganFranklin in McLean, Va. He recommends meeting with them to gather dashboard "options," including what data might be included and how it could be presented to different users based on their roles and responsibilities. That approach also can help with early identification of data quality and consistency issues -- and if data isn't reliable, users won't trust dashboards. "Bad data will kill adoption," Brydges said.
In many companies, the primary users of dashboards are senior executives who need to see only an overview of data, said Keith Metcalfe, a vice president at WCI Consulting in Plano, Texas. "Less is more when presenting high-level information," he added. "Don't get too detailed and make the dashboard screen too busy."
To get accurate analytical results and the most business value possible from dashboards, companies should develop common definitions for key performance indicators across different business units and departments, Metcalfe said. He also recommends keeping communication lines open with business users and setting realistic but aggressive timelines for developing and modifying dashboards. "The key is quickly getting a result to the end user," he said.
But that isn't the end of the dashboard design process by any means: Schiff warned that BI dashboards can get stale if they aren't refreshed to keep up with changing business needs and new data. "At a minimum," he said, "you should be looking at dashboards on an annual basis to make sure they're still actionable and relevant."
Alan R. Earls is a Boston-area freelance writer focused on business and technology.
Follow SearchBusinessAnalytics.com on Twitter: @BizAnalytics_TT