« Run That By Me Again? | Main | Wee hours in the Garden of Good and Evil »

February 6, 2008

Proof Politicizes Architecture

Management requires an ongoing accountability for effectiveness. Normally, this accountability is the recognized set of terms for "proof". The accountability includes some model of measures, and while the model may be good or bad, the agreement by interested parties to use the same model makes it the argument that counts the most, even if it is wrong.

The model then represents the ruling hypothesis about effectiveness and, in short, memorializes the belief system in place about how to make progress. This fuels, in turn, a prevailing culture of proof that casts its influence over the several necessary layers of enablement that allow for effectiveness to result. The problem is that this influence may be inappropriate, and to avoid being misled by it we need to know why.

The answer is that the layers of enablement are variables, and the combinations of their variations offers more than one path to effectiveness. To describe an occasion of how effectiveness resulted, we would monitor variations within threshholds on as many as seven layers:

1. Effectiveness is usually a measure of the degree to which an actual outcome matches a desired outcome.
2. But the outcome is a strength of reaction to an impact generated by some agent.
3. The level of impact observed can itself be more or less than expected, or desired, or needed -- and often there is insufficient attention paid to distinguishing the three characteristics (which is what allows the seeming inevitability of unintended consequences).
4. The cause of the observed level of impact is a state or an output.
5. The agent of that state or output is a quality level of process.
6. The runrate of the process consumes inputs.
7. The level of supply of inputs allows the processing.

In this view, the bottom four layers predispose whatever can happen in the top three, which is why the bottom four are so routinely captured by the design of an architecture. But if the architecture is unfamiliar or expensive, the distance between layer 4 and layer 3 is quite large. The burden of proof will fall to the architect to demonstrate bridging the gap.

With some "proof-of-concept", the effort is made, but it must be accompanied by its own "concept-of-proof" that relieves it from the pressure of "measuring up to" the top two layers.

It takes gutsy decision makers, or ones with little to lose, to accept this -- rather than to impose the top-line measure as the justification for considering the architecture at all. Innovators and early adopters accept their proof points at level 4, while laggards are hard pressed to dig below level 1. Consequently, the architecture in question gets positioned on the spectrum somewhere between demonized and championed.

Posted by Malcolm Ryder at February 6, 2008 10:37 AM

Trackback Pings

TrackBack URL for this entry:
http://www.malcolmryder.com/cgi-bin/mt-tb.cgi/351

Comments

Post a comment