Published in TDAN.com January 2007
Here is a quick question: at what point does the ability to analyze information for intelligence purposes overload the ability to exploit that information? An article recently published in the
Washington Post suggested that the human need to acquire additional data for analysis leads to a situation where decisions are indefinitely put on hold as more and more data is collected. In this
situation, people enter a “holding pattern” in which additional acquired information not only doesn’t add value, it begins to detract from the knowledge that has already been
We can probably assume that the same concept exists within the world of business intelligence – with the plethora of reports, scorecards, dashboards, widgets, etc. at our disposal, it would
be ingenuous to believe that individuals might not be overwhelmed by the availability of “analysis.” And despite the availability of all of these presentation mechanisms, we still
require subject matter experts to ponder and review these reports to provide recommendations intended to drive better business decisions.
On another topic: it is difficult to imagine traveling anywhere these days and being surprised at the ways that mobile technology has not only changed the way we do business, but literally changes
our perceptions of normalcy. Twenty years ago, a person walking around an airport terminal talking to himself would be considered insane. Today, almost the exact opposite is true – where
would we be without our mobile phones and Bluetooth earpieces?
With this in mind, it is actually interesting to consider how business intelligence seeps into the mobility layer. There is a great desire top be able to provide active reporting and status of
activities to individuals while they are away from there desks. It should be a natural fit, except for some small problems: the amount of screen space (“real estate”) and bandwidth
available is not suitable to deliver the presentation of knowledge in the manner to which end-clients have become accustomed. Trying to project a fancy graphics-laden web page onto a screen that is
2 by 3 inches is probably not a wise idea.
Clearly, one should not expect to be doing graphics-intensive analysis and financial projecting using a handheld device. On the other hand, what are people really looking for in operational
business intelligence reports? Probably one of a few things:
- Assurance that everything is continuing to operate as expected, or
- Notification that some exception event has occurred that requires attention, or
- Indication of some emergent business opportunity that requires attention.
In the first case, no action needs to be taken, but the other two cases do require attention. Basically, the subject matter expert scans the report (or dashboard, etc.) and takes action either when
something bad is happening or when something good is happening. If the process of determination of these situations can be modeled and automated, then in fact the business client doesn’t
really need to see the entire report. Instead, the client can be alerted to the situation with suggested actions to take. In other words, the approach to BI knowledge management becomes one of
“management by alert.” I have two recommendations and one caveat about this idea.
Recommendation number one: Limit the use of the mobility approach to communicating the results of performance and operational BI (reporting on operational productivity) along with knowledge
discovery (e.g., the results of predictive data mining). If it is true that the expectation of the user is to employ mobile BI for the purposes outlined above, then we really only require
notification of action events, relying on the absence of alerts to indicate normal operations. In turn, action events (either positive or negative) can be communicated simply, without having to
reproduce the entire set of presentations from which the conclusion is drawn. This provides the right kind of information without incurring the associated bandwidth or real estate overhead.
Recommendation number two: Develop the ability to integrate mobile BI delivery with real-time BI analytics. Consider the circumstances under which one might expect to make use of mobile BI. If you
are a manager who does a lot of traveling, while the allure of a mobile BI solution might be appealing, but if you are not in the position to take action, the extra information may be a hindrance
to competitive gain. However, if you are a floor manager for a factory, and require instantaneous alerts of productivity or danger conditions associated with operations, real-time reporting is
critical to business continuity and staff safety. Sales, supply-chain, and manufacturing are all examples of applications where individual actions can be taken by staff people deployed across a
wide area, and can all benefit from real-time BI alert delivery to the palm-top.
Lastly, the caveat: Control the urge to become overwhelmed by alerts. An alert is in interruption of “regularly-scheduled” management. Yet the alert also seems to demand immediate
action, which will preempt current tasks. When the stream of alerts becomes so heavy that all people are doing is reacting to alerts, it deviates from standard operating processes, and leads to
chaos and disarray. Therefore, before empowering individuals with hand-held BI, institute a program for “alert management” that enables qualification and prioritization of alerts,
identifying actions that can be automated, and limiting the number of alerts to those that have clearly defined return.
Copyright © 2007 Knowledge Integrity, Inc.