'

ROI in SOA: Illusion or beautiful noise?

An SOA may be part of an orchestra making beautiful music, but it may be hard to pick out what parts of the music are a result of the SOA. A while back, on separate occasions, I had the opportunity to speak with Bill Inmon and Dan Lindstedt,  both leading thinkers in the enterprise data warehouse space.

An SOA may be part of an orchestra making beautiful music, but it may be hard to pick out what parts of the music are a result of the SOA.

A while back, on separate occasions, I had the opportunity to speak with Bill Inmon and Dan Lindstedt,  both leading thinkers in the enterprise data warehouse space. Both raised an analogy that strikes me as apt for SOA projects as well.

That is, a well-run company is like an orchestra, with many parts coming together and performing well. The issue then, is, when a particular project, such as an SOA effort, is put in place, how do we determine how much ROI it is generating? How much is a direct result of the project itself, how much is the synergy generated with other projects, and how much comes in spite of the project?

A company with an ambitious, enterprise-wide SOA effort is likely to be a forward-thinking one to begin with. ROI -- measured in either additional revenue or cost cutting -- may come as a result of good management practices, with or without SOA.

As Dan Lindstedt put it (again, talking about enterprise data warehouse, but I think his words are just as applicable to an SOA): It's near impossible to discern how much ROI a project is contributing. Many organizations may be struggling with a large SOA project while "also struggling with other initiatives to clean up business processes.  So who's to say that the both of them together didn't help each other?  Who's to say that I can't attribute the additional revenue to a combination to the two projects succeeding? Who's to say that if one of those initiatives failed, what would have been the outcome?"

The other side of that is, one project may not have succeed without the other -- an SOA effort may have not succeeded if a business process improvement initiative didn't also succeed. Such projects are "linked together, and to pull them apart would be very challenging. In a symphony, if the the violins recognize the horns are playing particularly well tonight, what would the individuals in the violin section do? Since they hear the energy, and they hear the horns trying to do their best, the violins may step up to the plate and say, 'I'm going to play a little bit better tonight because this sounds so phenomenal."

Good, inspired management breeds success. The spirit of innovation plays just as much a role in the success of an important project such as SOA as any other variable. But it's often difficult to pinpoint the exact source of ROI.  It just all sounds great together.

Will it even be possible, then, to "hear" the part SOA is playing in your organizational symphony? Or is the sum of success more than its parts, be they SOA or other significant projects?