I did not need to do that a lot, so I have no idea how consistent the method is, but when I did it, I just used hot air to solder the pad and then drag soldered the rest of the pins. This let me check that the pas is actually soldered. The rest of the method is just just QFN - heat up the device, poke it a bit to make sure it settled, make sure pin alignment is still ok.
It was a bit annoying to manipulate the large package, since it did not fit entirely into the field of view of my microscope, but I managed to make it work in the end.