If you weren’t convinced that dashboards have become over-hyped, these two Dilbert strips that appeared last week should be the final straw:
(Click twice to enlarge the comics)
The intellectual side of me would like to defend our profession and chalk these up as simply cartoons designed to elicit a laugh. Unfortunately, they are accurate depictions of the reality in many organizations; the data within them are either inaccurate or simply not believable.
Dashboards need a certification process for all of the data they contain: goals, initiatives, financial and non-financial metrics. With certification and auditing comes trust. With trust, comes use. With increased use, more impact.
After all, that’s why we build dashboards. To have impact, to increase performance. Not just to have another IT project. Right?
(BTW, I waited until today to post these to see if Sunday’s strip – usually the most vicious and insightful – also was about dashboards.)
Manage By Walking Around 
Great thought on dash board. One of the best practices in this space is to show up the most relevant information that the user would want to see and make decisions on – and not just show up jazzy things wiht cool graphs that he or she is not going to use. The certification process is a cool thought but I would say that the goals, initiative should be integrated into a workflow process and therefore accepted. This could be achieved by integrating with planning system, where approval and sign up are important milestones
For very articulate blasting of cool graphs that have very little real life usefulness, check out http://perceptualedge.com/blog/
Our group has coined the term WOD for “Wall of Data”. Even when you’re tracking confidence levels in a metric, which is almost never done, but when it is done, 95% confidence is almost universally used. That’s 1 in 20. So tell me what happens when you create a WOD with 20 or more metrics in it? There be ghosts in that thar machine …
I’d be interested to know who assigns the confidence level to the metrics. If it’s the same person that creates/populates the metric, then I suspect the confidence value is practically meaningless. There has been lots of research which shows that the closer you are to a “fact”, the more confident you are of it. From my point of view, certification must be done by someone independent. Think of external auditors.
[…] transplant something that Jonathan wrote when discussing management dashboards, I think it is equally relevant […]
[…] references fellow SAPper Jonathan D. Becher who says: Dashboards need a certification process for all of the data they contain: goals, […]
Dear all,
I respect your thoughts but beginning of innovation is always in a crude form.
But by developing the innovation by supporting it with more innovative ideas will be more helpful for everybody rather than stoping it progress by shear criticism.
I am working on D/B and i know that if this concept is develop than this will help in reducing lots of time and human intervension in data interpretation.Any method which can save time must be given attention.I agre there are certain drawbacks but there can be ways to reduce or remove them.
Thankyou
Gaurav-
While I agree that one should give early innovations some lattitude, dashboards do not fit this bill. People have been creating dashboards to display metrics for at least 20 years.