Nuts and Bolts and ROI

I was asked recently to comment about the return on investment for service oriented architectures. Because SOA is the approach, not the solution, cost justifying it in the context of a quarterly budget is tough – and perhaps even counter-effective. I realize that this is not the best answer to folks who are trying to sell an investment in MITA to a CIO, but it is, unfortunately, the nature of the beast that ROI in today's IT market makes a direct correlation between the investment in the nuts and bolts of the system and performance measures of that system. In that light, an MMIS that is built using an SOA approach may provide a faster and more accurrate claims processing engine but this performance improvement in moving claims through the mill misses the mark in terms of the real value of MITA - maturing all of the business processes assciated with the Medicaid Enterprise and improving the quality of care being given to Medicaid beneficiaries. ROI in this context will ultimately be measured in the lower program costs that result from improved benficiary care. Because program dollars and IT dollars usually come out of different pots, it is difficult at best to correlate cost savings across the two budgets. Lower application maintenance costs are the closest thing I can come up with as a tangible IT ROI for an SOA build related to MITA. I am interested in hearing from anyone who has faced the challenge of selling MITA in terms of ROI to a State CIO.