Mitigating the risk of using semantics in the Enterprise

I was talking to a project manager within a large and very sophisticated enterprise this week about why her organization chose to go with an ERP-based implementation as opposed to a semantics-based implementation on a new master data system.  She was pushing for the semantic solution, but they weighed the pros and cons to each approach and came down in favor of the ERP system. 
Mostly it came down to “risk” mitigation.  Despite the cost and time projections for the semantic solution being lower than for the ERP one, and it even though it would have been more flexible, the semantic approach was held back by no internal use cases, limited technical expertise on staff, and a low degree of confidence in the cost estimates.  Certainty prevailed over uncertainty.
It’s been said many times before but we need more case studies of proven, successful applications based on semantic technologies, or else semantic technologies will not make much progress in the enterprise.   We have some published on the Semantic Universe site but I’m redoubling my efforts to identify and publish the success stories. 
Please contact me if you have a good story to tell, especially if you can define specific metrics of success in terms of resources saved, opportunities gained, or some other concrete ROI measure.