@User yes that sounds good to me. How feasible would it be to get the backannotation script to figure out what the resulting instance name in the netlist is from the instance name on the the schematic? So you can for example say m4 (or even xm4 since we know what the spice instance name is at placement) and the script could expand that to the full name. I guess you'd have to make some assumptions about the subcircuit now I think about it. What do you think?