--- In [email protected], "Anne Thomas Manes" <[EMAIL PROTECTED]> wrote: > > +1. > > To rephrase, lack of governance (i.e., the complete process) > severely impedes SOA. > > Jeff asked me this question privately last week. Today he asked me > whether 'business & I.T. alignment' is killing SOA. My response to > him was: > > I think lack of understanding of what it really takes to do SOA is > what kills SOA.
Could "SOA" in the above be replaced with "architecture" (of any sort) and the statements still hold true? IMO, yes. It's not the SO part that is the issue. It's the A part. > To fully deliver the returns promised by SOA, it > needs to be part of a much larger initiative to transform IT. The > following perspectives will significantly limit the potential > return on investment: > > - Technology-centric effort > - Project-centric effort > - Integration-focused effort > > Characteristics that facilitate success: > > - Investment in social capital, i.e. learn to speak "business" and > establish trust > - Enterprise-wide perspective > - Prioritization of projects based on desired business outcomes > - Focus on improving data quality and process optimization > - Major changes to the IT department, e.g., new CIO, reorganization, > adoption of agile > > Anne > > p.s. (this was not part of my original response to Jeff) > The major changes to the IT department aren't essential, but they > cause a massive shake-up that typically sparks a significant change > in the attitude of the IT staff -- both in the way they interact > with each other and the way they interact with people in other > groups. It's the change in attitude that facilitates success None of these characteristics (the wrong perspectives nor the keys to success) have anything to do with SOA, IMO. They are important for any enterprise level architecture to succeed. Indeed, they don't apply only to architecture efforts either. They speak to corporate culture and relationships among groups. They are aspects that have a broader reach than just architecture. Which is of course your point (I think). Please correct me if I'm interpreting your post incorrectly! So those that are brought in to a company to help "do SOA", (such as the fine folks at Burton Group!) end up needing to "fix" the company interaction dynamics first in order to succeed with that effort at the enterprise level. Is that a fair statement? It seems that the prerequisite characteristics are independent of architectural style. So would tying those characteristics to SOA be inappropriate? Is the focus on "transforming IT" misplaced? Perhaps that would be better viewed as a potential byproduct of pursuing the characteristics above--which are intended to for an end goal of a more successful business rather than a goal of "transforming IT." Aren't we trying to transform/improve the business? -Rob
