Exactly—scalable, the same without having to go back, but more flexible; rather than winner takes all, I'd have several. If you're doing workstation support, for example, have several vendors involved so you have the ability to move work back and forth, depending on performance. I would try to have a bigger ecosystem of suppliers.
I tried very hard to get government to continue to invest in its ongoing institutionalization. They chose not to, at the end of the day, and I think they're paying for that decision, because they have to restart everything now. Make sure you're taking that knowledge and embedding it inside a government over time. Working on that stream is critically important.
If I were going to look at the federal government now, and I think it's an enormous opportunity, I wouldn't try to replicate it. I would still start with a central group. Defence might have its own just because of its size, and then everyone else would have a centralized group.
There's a co-accountability between the leader of that group and whatever deputy minister's running a project. If it's health, if it's tax, whatever, it's a dual accountability, so that both parties are dependent on each other for delivering it and getting it done.
I would start with one central group for most of core government and I would make its use mandatory and I would give them veto power over the enterprise, because if it becomes voluntary, it won't get used. They need to be funded; they need to be free for the departments so that they're adding value to those departments, but the departments shouldn't have to be taxed to do it right. Then over time you become greatly in demand, but at the beginning, I wouldn't make it a tax on projects. I would fund it centrally, but I would make it mandatory.