Business Intelligence in general, and Executive Dashboards for Business Intelligence specifically, are current areas of extensive blogging, lots of development activity…and in fact executives are making alot of money selling these dashboards to other executives. Unfortunately, it has also resulted in lots of incorrect marketing claims, urban legends of how executives think and what they need, and bluntly lots of very poor decision support design masquerading as ‘dashboards’. These (incorrect) urban legends have become a sort of societal meme, becoming widely held beliefs, but nonetheless incorrect. (Where is Snopes.com when you *really* need it?) The Human Factors and Ergonomics Society is even accepting ‘research studies’ on whether adding a 3D drop shadow behind a Red-Green-Amber coded tachometer degrades an executive’s ability to judge ‘how far into the red are we?’ (See for example Chipley, Barrow 2007)
Some of these common misconceptions of executives and their dashboards include:
- Executives are very busy and therefore can’t be bothered with ‘the data’
- Executives like/prefer/demand Red-Green-Amber stoplight charts and histograms and pie charts, and…
- Executives can’t understand anything more complex than <that pitiful list above>
The current ’state of the market’ in executive dashboard design overlooks even the most foundational research into what constitutes expert decision making in complex work domains. (An excellent starting point is still H. Simon’s “Models of Thought“, where Simon characterizes chess experts as thinking in abstractions of chess while novices focus on exhaustive memorization of the physical pieces.) An effective Decision Support System should support–and even promote–such expert levels of decision making performance. Yet, imagine an ‘Executive Dashboard for Chess’, designed using today’s state of the practice R-G-A, histograms, and trend lines. It’s ridiculous to envision good chess play from a set of even many bars and needles indicating “chess metric 1-green” and “chess metric 2-yellow”, etc. (Remember you cant have TOO many metrics, executives will get overwhelmed, and it must fit on an 800×600 display, or better yet fit on the BlackBerry Bold’s 480×320). In short, even for an “enterprise” as constrained as a game of chess, such a ’state of the market’ Executive Dashboard offers no support for any abstract conceptualization of chess.
Fortunately, some of the ‘executives’ have learned exactly this lesson (the hard way) and have begun seeking the kind of powerful decision support they deserve. These executives are supervising enterprises with more moving parts than the Space Shuttle Program, enough employees/subcontractors/consultants to swamp several football stadiums, and often more riding in the balance than a balance sheet. Dashboards are beginning to be seen for what they really are: Directed Attention Longshots (see Woods, Hollnagel 2006) that must provide full-breadth, informed support to the executives’ initial focusing and refocusing for decision making. In many recent system design reviews, executives have begun to ask the right ‘gut check’ questions of their system developers, e.g. “But for an enterprise of my size, something is *always* broken/out of alignment/out of tolerance–this display will *always* indicate “in the yellow” which isnt helpful to me at all!”
Executives, as customers of their own decision support environments, have properly begun to demand Decision Centered Design (provided by Cognitive Systems Engineering methodologies) to deliver more powerful and effective systems and not just ‘dumbed down dashboards.”
Stumble it!