|
Enterprise Architecture Is Built For An Organization’s Culture
By Mike Kavis
Expert Author
Article Date: 2009-08-10
I participated in an excellent conversation today on Twitter between a few EA advocates. I was following a conversation passively until I came across this tweet:
Reducing complexity implies that we know how to measure complexity. How do you measure?
Then I joined the conversation and replied
Disagree, measuring can add complexity. Only need to measure if you need to sell the value of it.
Then the conversation really took off and went on and on for a while. We are all huge advocates of EA but have slightly different opinions about metrics, EA goals, and many other aspects of EA. The bottom line is that there is no one answer or one solution for "doing EA right". At the end of the day it all comes down to an organization's culture, the amount of influence the EA sponsor has, and the quality of people that deal with EA (directly and indirectly). But what triggered this blog post is when I started hearing language that talked about what EAs "must do" or when I heard generalizations like this - " I think we can do everything poorly or one thing well.". To me that is the all or nothing "purist" view also known to me as EA Theory. It is one thing to visualize a perfect world where EAs follow the plan to the very last defined process but the reality is that EA on paper is a much different animal than EA in the real world.
To put it in perspective, EA is like a marriage (I'll use my own personal example so I don't offend anyone). Before getting married, both my wife to be and I had a vision of what a perfect marriage would be like. Needless to say, hers did not include me continuing to play softball 3 days a week, fishing on Saturday, and watching football all Sunday (the "Purist" vision). So I knew going in that some sacrifices needed to be made and meeting half way would lead to a long lasting and fulfilling partnership (15 years so far). However, every person is different and had I chosen another partner, the amount of compromise might have needed to be more or less. The same goes for EA. If EA types go in expecting the "purist" view or the perfect world, they are in for an unhappy relationship. Just like my wife doesn't understand why my week is ruined whenever the NY Giants lose a game, the business won't understand why an architect gets ticked off when they can't measure all of the things they feel is required to measure. The bottom line is, you may have to give up some things in return for the privilege of having your EA initiative funded.
Demotivationals
Let me share a real life example of a failure that I was unfortunately involved with many years ago where the leader of a CMM initiative did not balance theory vs. reality. I was at a 100 person IT shop that focused on building software to deliver services in the form of analytics. We were a young company with very little process and were working hard but not smart. It was time to implement some standard processes. We brought in a new VP to lead the charge. This person had a vision of the future state which was to be CMM level 3 by some chosen date in the future and expected to be able to follow CMM by the book. The problem this person ran into was that the culture was not ready to go from chaos to level 3 overnight. Going from no peer reviews to having an army of people reviewing your stuff is a bad plan. There has to be a transition. Needless to say, this initiative failed miserably because the leader had a "purist" view which led to unrealistic expectations. There was no compromise. This marriage was doomed before it started!
That is what I think about when people tell me you must measure this and that. Maybe I should measure it but maybe the organization is not ready for that yet. It is not black and white. Keep in mind, that small and medium size companies also practice EA and resources and funding are typically tough to come by. My experience from spending 20+ years with small and medium sized companies is that the EA teams are typically very small and have limited bandwidth. EA theory is not reality in my environment. So we typically scale down or adopt portions of our favorite frameworks and make due with what we can. It may not make the purist happy and they may not even acknowledge that we are doing EA, but at the end of the day, applying any EA best practices is better than applying none.
In summary, I enjoyed the Twitter conversation today. It was thought provoking and everybody had good points. I closed with these Tweets before I had to leave for a meeting:
Sounds like a purist point of view. Goal should be improve incrementally each year. May never get to nirvana.
I agree with that. I am a huge advocate of EA, but I have seen EA purists derail efforts by not balancing theory vs reality
So I am not bashing anybody's views here and I am not saying that we shouldn't measure things. What I am saying is that every organization is different and therefore every Enterprise Architecture should be expected to be different as well. That is the difference between EA Theory and EA reality.
Comments
About the Author:
Mike Kavis is a veteran Chief Architect with over 23 years of IT experience including distributed computing, SOA, BPM, data warehouse, business intelligence, and enterprise architecture. Read Mike's blog at Enterprise Initiatives.
|
|
