The Political Incorrect: Drop Croudsourcing
Croudsourcing is on eof the many web 2.0 hypes with overinfalted expectations.
You might get more inouts and between the many more inputs will of course be good stuff. However, crouds are terrible and often even negative intelligence in evaluating suggestions.
Croud analysis CAN be good when the problem is welldefined, but this problem is complex, cross-competeance, poorly understood in almost every dimension and overloaded with ideological assumptions and interests.
I would suggest a return to rationality and arguments instead of a process dominated by populism without linkage to causality and risktaking.
In computer science there is a concept called late binding meaning you excplicitly delay impiortant choices that cannot be finalised with existing knowledige.
Innovation is about continous change - any legacy creating choice may block an important innnovation. And public sector ICT is almost always characterised by a lack of focus on continous change and runtime demand-driven choices.
The best public sector solution is one that from day one consider itself obsolete and ready for replacement.
Heike Gerken commented
Crowd! Kraut! Cloud!
but not Croud...