Just an idea, can't you just swap staging and production? So, actually the system you've tested goes live by switching nothing more than a pointer (no deploymen

Just an idea, can't you just swap staging and production? So, actually the system you've tested goes live by switching nothing more than a pointer (no deployment involved). Won't raising support cost at some point suggest it's cheaper having two swappable live systems than the alternative?