« Solution Strategies: Solving the Right Problem | Main | Be careful what you ask for: you might get it »

August 31, 2005

When Measurement doesn't help Management

Organizations increasingly manage the support, timing and priority of activities by using metrics. And the activity called "measurement" should not exempt from being managed, so naturally there should be measurements about measurement. This might be a hidden dimension of "management".

If we can't manage what we can't measure, and thus if measurement is a basic component of management, then the performance of our management effort is dependent on the performance of our measuring effort. But how dependent?

In what follows, let's look at how and why measurement is:
- critical but can still fail to support good management; and,
- potentially strategic to good management, if measurement itself is properly managed.

Measurement creates the formal definitions that will be used to identify and track states and events in operations and in the operating environment. So, what we come to "know" about our conditions can be deeply tied to the way our measurement allows us to recognize things.

In the classic hierarchy of data leading to information and information leading to knowledge, definitions generate data, or "facts", that are interpreted by a perspective into information, but information is not knowledge until it is assigned to a class of concepts. (This perspective-driven assignment of information is what we call "context".) Meanwhile, the definition of the concepts is not dependent on metrics, but rather on classification -- a different discipline.

Concepts is where management first stands beyond measurement. At this point, though, what management gains from measurement is the engineering, refinement and/or updating of prior knowledge into current knowledge.

The value of this knowledge then begins with applying it. The two main ways of applying the knowledge are:
- to distribute it; and,
- to use it as a decision-point.

For the purpose of "management", there are two distinguishing keys to applying the knowledge.

The first key is to remember that it is not "information" being applied, but "concepts". We're properly very leery of trying to manage without information; but measurement is only one source of information, and regardless of the source, information is not meaningful without conceptual associations. When information is received and processed, what should be further distributed for management's sake is the concepts. That is, in the next step, communication should drive the simultaneous, interlinked distribution of both the information and the meaning.

The second key to the management-oriented application of knowledge is to see that action is another way of propagating concepts. When actions are classified -- by purpose, impact, and propriety -- they can represent a model of operations that describes how the organization qualitatively distinguishes itself from others. Decisions typically select preferences that in turn select and promote actions. This is what "management" does with knowledge that drives opertional performance. Additionally, it reinforces or proposes the prevailing desired identity of the organization.

Because of what it represents, proposed or dictated action becomes a form of communication in itself. Starting with decisions, management must do a good job of forwarding both the ideas and the actions that matter.

This forwarding or promotion is the second point at which management stands beyond measurement. For many reasons, managers promote things according to preferences and tolerances, not just neutrally or automatically. Whether the driving reasons are psychological, political, tactical, or whatever, promotion can run independently of measurement.

However, when the promoted actions and ideas are then monitored and measured for their timing, compliance, impact, etc., management comes full circle as there will be new data generated about resulting states and events. Recognizing states and events can be facilitated by using the language of measurements, but not all states and events will have already been discovered before: unprecedented effects will occur and will need a way to be identified beyond the current terms of measurement. Research on possible effects should extend beyond the local or immediate management milieu, and measurements should be upgraded to accommodate the newly acknowledged effects.

In review of the above, looking at how definitions, concepts and promotion line up offers the picture below:

How good is your management at getting all the way around the circle? This question asks about the performance of management itself. Seeing that there are at least three major components in the cycle highlights the idea that measurement will be a prerequisite of good management but not necessarily a cause of good management.

The ideal situation would be that measurement is refined to the point where it is not just critical to management but actually strategic. This would mean that measurement should have both methodologies and deliverables that actually make management's overall effort the differentiating factor in generating the value of what is being managed.

We know that poor measurement can be a significant inhibitor of good management, because measurement is critical -- that is, it is a prerequisite whether it is helping or not.

By explicitly addressing tie-ins with knowledge and communications, measurement stages an opportunity to become strategic and not just critical.

Note: The discussion above focusses on characterizing the relationship of three types of "business information" (measurements, knowledge and communications) in a singular management lifecycle. One online resource of many in-depth papers more conventionally discussing measurement is the "TechRepublic" free-membership IT library. A good current example of the available content offered is the whitepaper The Metrics of IT: Management by Measurement from the Enterprise Computing Institute -- reached via TechRepublic membership at this URL: http://techrepublic.com.com/i/tr/downloads/home/metrics_of_it.pdf

Posted by Malcolm Ryder at August 31, 2005 6:54 AM

Trackback Pings

TrackBack URL for this entry:
http://www.malcolmryder.com/cgi-bin/mt-tb.cgi/118

Comments

Post a comment