Saturday, February 16, 2013

Is Dashboard Out of Date

“Can the same image prompt different stories and memories in different people? That’s a good test for a “super-graphic.”   -- Edward Tufte

Dashboards can be extremely useful in a lot of areas, whether it’s a decision-making tool for executive, sharing information to an employee or supplier, or analyzing costs per sector. It’s a higher graphical representation that many people feel comfortable using. However, many of today’s dashboards are not very effective for various users to capture business insight, so is the dashboard out of date?

1.    Causes of Ineffective Dashboard 

BI and Data Analytics are hot deployments in businesses today; however, are the dashboards smarter than twenty years ago? Do they prioritize information enough? Are they action-based?

  • Ineffective dashboards are the result of developers’ failure to intimately understand the business model and decision-making process. Thus, it is not capturing the essence of the business in the graphical interactive interface.  
  • An effective dashboard is harder to design than it looks. Design and Development are two distinct skills and you can't rely only upon programmers to make insightful dashboards and give them exclusive ownership. Everyone has to have skin in the game to make it work. A good developer is not necessarily a good dashboard designer who not only must invest some time to understand the business model and decision-making process but also have a good grasp of visualization, which implies a good understanding of human visual perception and cognition. 
  • It could be data or quality issues or people are not trained to produce the results with the data so that any presentation visual or not wouldn't be effective, garbage in, garbage out. However, "low quality" speedometers gadgets could use a facelift, or perhaps they need more vertical market designs that capture the essence of the application. 
  • Fancy doesn’t mean updated or effective: Many fancy dashboards, maps, or policy visualizations aren’t designed with the end-users in mind and won’t help executives answer the most basic questions. If you had a situation where the organizational response to some "dial movement" was obvious, then there's little point in presenting that information via a Dashboard and waiting for some bored executive to notice.

2. How would a Dashboard's Functionality and Performance be Assessed?

The Purpose of Dashboards is to provide visual, summarized presentations of whether "results" are being achieved/trending as expected, or not. Where deviations from expected seem material, then some level of analysis to discover the causes would generally be started, and then based on that analysis, some putative cure would be put into action. Depending on the complexity of the situation, that could take minutes, hours, or months. 

  • Categorize the different types of reporting dashboards, but should classification be based on decision-time horizon since "actionable" is the debated word?
    1) monitoring (real-time)
    2) operational (tactical)
    3) performance & financial (strategic) 
Eight functional points to evaluate a dashboards:
1) interactivity
2) visualization/graphics
3) drill-down capability
4) simplicity
5) point of view analysis
6) navigation
7) actionable 
8) User intuitive

3.   Future Trends of Dashboard and Best Practices 

Perhaps as dashboards presentation formats move from being mere generic BI tools with Big Data/dashboard functionality to vertical market Big Data solutions, they will start to capture "the essence of the business” in the graphical interactive interface to allow data blending and have fast in-memory data processing.

  • Mobile Trend: The dashboard data visualization format is all the more important as mobile devices "re-educate" business managers that there is a better way to consume information other than standard column/tabular Excel type of reports. It's a mind-shift. "Telling Stories with Data." 
  • Climb Maslow Pyramid:
1) Start from basic: Before running, one needs to learn how to walk! Strange enough, most organizations still don't have easy, user-friendly access to fresh broad data so the first few dashboards required are not rocket science, and don't need to be! 
2) Richer Dashboards: Later on, when the first layer of the pyramid is solid, then it is possible to start talking about richer Dashboards that help to go further.

  • Interesting and related to Dashboard "best-practices" or statistics:   
1) Collect the KPI from the business and design them so that they are personally connected to it and draw some information out of it.
2) Cultivate designers’ capability to understand the business,  dig into the business problems, and to propose great solutions based on his/her knowledge of the various type of analysis available.
3) Advanced data visualization is based on the fact that 70% of the human sensory receptors are dedicated to vision while the other four senses share the remaining 30%. 
      4) Human brains are much more effective in recognizing shapes trends and
          patterns than analyzing spreadsheets or tables full of numbers.
     5) Visual data analysis principles are based on the work of Edward Tufte and other
          visualization expert.       

Seeing is believing! An effective dashboard is like a widow, helps capture business views and see-through business insight. “The best design gets out of the way between the viewer’s brain and the content  - Edward Tufte


Post a Comment