As governments embrace service-oriented and performance-based procurement, they are discovering one major problem: they are not very good at it. And this has serious implications for the adoption of open source solutions by the public sector.
Performance-based procurement of services presents new challenges, as U.S. federal agencies are learning. Services are not managed, measured or maintained in the same way as products. The focus is service needs, not products specifications. Procurement personnel trained in buying hardware and systems are spec-oriented, not outcome-oriented.
These difficulties are compounded with open source. Open source is just as disruptive of government procurement as it is for the IT industry. Worse, open source involves the procurment of both a software product (community-developed and often free) and support services (with performance-based contracts).
Common criteria for bid selection -- how well product meets specs, financial stability of the company, product costs, and alignment with current suppliers -- make no sense with open source. Getting the software and getting the support are often two different things.
While open source fits well with service-oriented procurement and its performance-based contracts, governments so far do not. More is required of IT staff with both open source and performance-based contracts before procurement begins. They need to understand the open source product/services and the outcomes they expect. They will need training to handle new demands with respect to estimating contract costs, conducting risk assessment, setting realistic baselines, performance benchmarks and financial incentives.
Metrics must become their new mantra.
Without a serious commitment to training, it is a lot to ask of any Acquisition Department.