I should preface this post by saying that I am a huge fan of enterprise architecture and have been for some time. However, I posit that there is simply no evidence that enterprise architecture actually exists or will ever exist.
If EA were working or was economically viable we could have expected a decline in the global outsourcing market as businesses sought to integrate and consolidate applications along an enterprise metamodel. However, this work predicates a level of control and ownership which is not consistent with applications/business process outsourcing.
Definitions of EA aside, it seems clear to me that the economic benefits of EA simply cannot compete with the economic benefits of outsourcing. In order to turn the tide, my personal opinion is that EAs need to be able to architect from the financial instead of the minutiae of the technical. Stop trying to solve technical problems and start trying to solve business problems. More importanlty – in line with the ‘E’ in enterprise – be able to cut through the petty bureaucracies and fiefdoms of middle management and begin to define business problems by calibrating financial statements with technical information. In other words, how can technical architects analyse adverse financial variances on the monthly balance sheet and turn them into changes in the technical development or delivery? What is the technical implication of a month-on-month 17.5% adverse variance on Indirect Labour? Until these figures become the mainstay of an Enterprise Architect’s work then EA will never be able to compete with the simple value proposition of outsourcing, i.e. “it’s not working! give it to an expert.”
More importantly, in virtually all businesses (i.e. non-tech businesses) the company is defined by accountants and sales people. Operations comes in close second and IT is so far at the back of the corporate parade that it can’t even hear the marching band play. In order for Enterprise Architecture to morph into something useful – and something closer to what I imagine Zachmann imagined – it will need to deal with enterprise engineering. This is the difference between systems engineering and Engineering Systems, i.e. the engineering of information between the various systems/functions in the business. What is the atomic meaning of a cost variance in IT terms?
Until EA/EE can truly grasp the complexities of the enterprise it will only ever be IT window-dressing. In the meantime, I hope I’m wrong.