I think it's very much based on the relationship between a service provider, in this case SSC, and the people who are receiving that service. Not everything is going to run perfectly every night. It's just not. The strength of the service element that you develop as part of your shared services organization is critical. There are so many moving parts, and some of those moving parts will go wrong on a pretty constant basis.
On the process side, Grant mentioned the ITIL world, the standards around how you implement service within an SSC. That is critical, because you need to have process to fall back on. You can't be fighting fires on an ad hoc basis, on a daily basis. You just can't do it. On top of that, you really need to have that incredible relationship between whoever is in a position of responsibility for SSC and those CIOs. You live or die by that.