The link between strategy and standards

I blogged a little while back that there is some interest in creating a common naming convention for enterprise web services within the company's IT group.  I've been looking in to this.  One thing that came up: if we don't have consensus on how we tie the EAI strategy to business goals, why should we proceed on a standard? 

The notion is: if we proceed without a solid connection to business goals, we may lead people in a direction that is ineffective or just plain wrong.  On the other hand, if we wait for strategy to produce basic standards, we will fail to lay good groundwork.  When we do decide to proceed with a common EAI strategy (which is a foregone conclusion), then we will be further behind that we should be. This would slow adoption of common integration patterns across the board.

Don't get me wrong.  There is a huge amount of integration going on.  However, different strategies are being followed by different groups, and there are literally thousands of home-grown applications being used to run the business.  (That's the problem with having too many developers!) 

It is my supposition that we could benefit from creating a common strategy for EAI integration that we can evangelize throughout the company's IT division.  I have very strong ideas about how this should look, but I'm not the first, and I'm will not be the last.  Common strategy requires intelligent consensus.  And it will not be free.

And there is the crux of the issue.  If we spend money, are we doing it wisely?  If so, we should see a business benefit.  In order to understand and measure that benefit, we need to tie to a beneficial business strategy.  In other words, we need to get our ducks in a row and describe the ROI of a common EAI strategy.

This is coming.  We have a terrific team.  I am truly excited about it.

So a standard is being created in lieu of a tie to the business strategy, but not as anything enforcable.  Rather, as a de-facto standard that we can share and begin to use, but which we cannot enforce or necessarily encourage until we have some way of knowing if it is the right direction to take.

And that, folks, is enterprise architecture.