The need for enterprise architecture
This post is also available in: Nederlands (Dutch)
Society is undergoing fundamental changes.
This is not unique for our time and age, but both speed and depth of these changes are probably greater than we have seen in centuries.
Enterprises are, whether they want it or not, both party and subject in these changes.
The way businesses were organised in the past does not scale into this age, and certainly not in the “future” age. Wrestling with larger concerns, in numbers of customers (for example in the case of producers of software), in chained businesses (for example in the energy sector), or in the number of anonymous shareholders and the incomprehensible speed of the trade in those shares, these companies are looking for their new identities to survive, let alone prosper.
Enterprise architecture (hereafter shortened to EA) is both prospering and struggling. It is prospering for a select group of practitioners and enterprises who have seen the advantages that can be derived from it. We can see this reflected in the vastly growing interest in EA frameworks like TOGAF or Dragon1 or several others. It is struggling for the majority of enterprises who have maybe attempted one or more EA implementations that failed horribly by consuming resources of time and money without returning any (visible) value.
There is (still) a dire shortage of experienced and effective enterprise architects. The use of the frameworks mentioned above is mostly still in its infancy, and most enterprise architects I know working in these disciplines are struggling within the isolation of their own organizations, not communicating and interacting with their peers.
In spite of these considerations I must still defend the importance and role of EA. Well, I am an enterprise architect so that might not surprise you. But my field of expertise is sufficiently broad that I think I can claim the seniority to view the issue relatively unhindered by personal bias in this.
The reason I stress the importance of enterprise architecture this way is that I think that within the unfolding complexity we have no alternatives. EA offers the only way to condense knowledge and wisdom, and managing tools, to deal with this. We still haven’t seen enough tooling that is able to deal with a sufficient level of realism with things like stochastics, uncertainty, change, waiting lines and complexity (in its scientific definition). But these tools will hopefully mature soon. The frameworks as well still focus mainly on helping architects to get a grip on the structure of enterprises. What is needed to become more effective is tools to deal with the dynamics of enterprises, the way they change and visualising these changes with a set of robust simulations and what-if scenarios. Something that needs to originate from EA.
With dynamics I do not mean the processes, because I categorise these under the structure of enterprises. I am talking about changing those processes, about playing with them, and keeling their orientation towards demand-chained and non-deterministic processes.
With an EA empowered by those tools we can move forward and discover what enterprises will accomplish then.