Yes, I will try. Let me give you a very old example that actually ended in a performance-based process.
In the 1980s, we were trying to buy a low-level air defence system for our forces in Germany. Our engineers had tried to write a specification on how to build an air defence system of missiles, guns, command and control, and communications--with no success.
There were many firms in the world that delivered air defence missiles, guns, and so on.
In 1985, a new project manager said, let's just ask industry to propose how they would defend a brigade in Germany. We went out for the first time ever—and I was a very young major, posted for the first time to international defence headquarters—and stated the operational requirement: to shoot down how many aircraft in how many minutes, day and night, in the German brigade area.
We received 13 comprehensive proposals from industry. They evaluated those primarily on their operational effectiveness, on their industrial regional benefits, on price, and on in-service support. The operational effect was weighted higher than anything else.
That entire process, from the point when we left the technical base and the stage when we'd tried to design it ourselves, which had gone on for about eight years, to the performance-based one, happened in two years, and we had deliveries a year after that.
In my view, that system is still today one of the best in the world. We have maintenance problems now after 20 years, but it does work.
We didn't do it again after the low-level defence in the intervening 20 years. We are doing it now for tactical airlift.
We did it with Nyala to buy the armoured vehicles for Afghanistan, for example. We went out and said, these are our high-level requirements; we do need them urgently.
So the delivery schedule was a key performance requirement. In fact, we only had one company, the South African company, that had any in production. There was just nothing else in production.
We delivered 75 Nyala in one year, and we saved our soldiers' lives.
There are lots of examples. There is one good old example, and I think there are lots of good new examples coming.