You can measure the gm of you input transconductance stage.
Typically for a Gilbert this is a source-coupled-pair or emitter-coupled pair. The gm reported by the DC operating point is usually a good start. For hand calculations you look at something like Gray and Meyer.
If your Gilbert transconducatnce stage uses local series feedback (i.e degeneration) then the gm becomes gm/(1+gm*Zdeg). Again this is covered in Gray & Meyer and other standard analog IC design texts.
Anyway irrespective of linearity technique/input toplogy, I often just DC sweep the output differential drain current versus input differential volatge. The derivative of the current with respect to the input voltage gives gm. Its a quick look see and is often sufficient if devices are operating upto a tenth of ft.