« Why performance evaluation must precede cost reduction | Main | Understanding the Value of Quality »

July 28, 2005

Managing without Metrics?

Metrics mania is for real, and not having it could be more of a problem than the problem we have with surviving the mania. Why? Most managers would not be able to communicate effectively without metrics, and with the pace of change in business, new metrics are necessary. The intense community-wide pursuit of more metrics means that there is always a better chance that the new ones you need are already discovered and available.

But management didn't grow from metrics; management adopted metrics. There is something deeper than metrics that characterises true management.

What is it?

Here is the really big notion that always drives my own sense of what "management" means.

Start with the idea that there is initially an item of some kind (an organism, process, object, etc.) going through changes and/or behaving in various ways, independently of your influence on it. Then, as soon as you have a particular objective for the behavior or changes of that item, you begin attempting to exert some logical influence on the changes or behavior to cause it to conform to your objective.

But you don't just take your best shot at that and see what happens.

You sustain attention and influence on the item's changes or behavior as much as is necessary to maintain and/or improve its conformity to your objective. This effort includes the possibility that you will at some point change your objective but attach it to the same item as before. Practically speaking, that may cause you to change the specifics of how you influence that item's changes and/or behavior.

As a result, you'll be busy with determining why things you've been doing are working well or why they are not -- and that is a two-way street: how manageable is something, and just how is it manageable?

The key to succeeding in your influence is to identify the characteristics, attributes, properties, etc. of the item that you can "engage" with the characteristics, attributes, properties, etc. of your mode of influence, such that your influence is *at least virtually* causal towards your objective.

I say "virtually" only because in practice the success of "the influence" is sometimes hard to confirm as being a coincidence versus a direct effect. I usually define that more precisely -- as confirming whether something is a prerequisite (exhibited by correlation) versus a cause (exhibited by testing). A prerequisite allows something else but does not necessarily cause it.

But here you can see why management is essentially dependent on a logic. The whole idea of management assumes that your influence on something is effective because it is "logical".

This means, furthermore, that obsessively measuring stuff outside of a logic of action to influence progress towards a designated objective (other than measurement itself) is NOT "management".

At the same time, the notion of "measurement" needs to be understood broadly. Here, the working definition of measurement is:

the determination of a significant distinction,
with the type of distinction being already named so that
comparison is straightforward.

That is, measurement determines that:
something is, to a discernable degree, of given state X,
and is therefore not of some other degree or some other state.

In the practices and measurements of a particular field, what becomes really crucial to management is to decide the reasons and standards for using the "degrees" and "states" being researched, referenced or communicated. From that, a logic of influence can be agreed and developed, and measurements appropriate to the logic can be cultivated.

Posted by Malcolm Ryder at July 28, 2005 6:46 PM

Trackback Pings

TrackBack URL for this entry:
http://www.malcolmryder.com/cgi-bin/mt-tb.cgi/87

Comments

Post a comment