Analytics header

Thursday, January 1, 2009

What to do with the politics in systems implementation projects

In the past few months I have been involved in software implementation projects for some big life science companies. It has been a while since I did this type of role so it has been somewhat of an experience getting back in the swing of things. The fact that I have returned to what was once familiar territory gave me a chance to reflect on some peculiarities and phenomena in these types of engagements, one of which as everybody surely knows is politics. In every such situation I really try to understand the pros and cons as well as the causes of these phenomena. 

Let me start with the good things. In a way politics forces the people involved to seek consensus. This process involves a lot of discussion in which the solution and the problems are reflected on with many people providing input. Ultimately this can lead to, in certain situations, to alternative and possibly better solutions. Another result of these “political” discussions are of course the fact that you get to dig into the detail and thus uncover certain aspects of the problems that would have been found later in the process, at a time where design decisions are more costly to undo.
When working in such environment I find that everybody is trying to get as much detail covered as possible. There is a feeling that we need to get all the details hashed out before we move on. Obviously this is result of our “systems theory” upbringing. By that I mean that the way most people, with technical backgrounds, learned to tackle complex problems is by a process of functional decomposition in which the problem is taken apart to an atomic level in the analysis process. 

The underlying assumption for this approach is that everything is known and that from the time the decomposition is performed the “system” will never change again. Both of these assumptions are of course false. We will never be able to uncover all the required detail and once we have most probably the details are irrelevant because the circumstances have changed.

The approach that I usually take is grounded in some of my earlier work understanding complex adaptive systems and Holonic systems or in other words systems that are in chaos. I will leave the detail out at this point and you can read more in some of the other posts on this topic. The bottom line is that in order to be most efficient one has to strive to identify recognizable patterns that may indicate possible problems and solutions. This of course takes a lot of experience and becomes kind of an art to know exactly when enough detail has been gathered to move on. However it is not "rocket science" nor does it require any special natural abilities. All that is required is the ability to embrace a radically different approach at problem analysis. In a way one has to learn learn new tricks -  which of course may be hard for old dogs.

So back to politics…. Obviously the bad part of a politicized process is that it is inefficient, and much time is spent dancing around making sure that everybody is ok with the compromise that regardless of what people tell you will always be part of the solution. Also a lot of time is spent in typical CYA maneuvers, which is also something that will never be admitted to, yet it is human nature. I guess some can draw a parallel to survival instincts. 

The art in of all of this is achieving a balance that like most things in life is important. The approach has to be holistic in that the project has to be embraced with not only the technical detail but an understanding of the social aspects of the project. Looking at implementation projects in this way is no different than a study in social interaction. A balance has to be achieved between autonomy of the individuals, manifested by their personal motivation and goals (political or otherwise) and the goals of the project to effectively provide a system solution. By the way this is exactly the definition of a Holonic system, which leads me back to art of identifying patterns and using them for problem definition and solution design, rather than the traditional functional decomposition approach. On an end note, politics will always be part of any project, it is inherently how we interact in a social system, and it is not necessarily all bad if you understand the causes and use these to your advantage.