Although every Functional Decomposer program from the literature uses certain strategy for finding bound and free sets of variables (variable partitioning), and/or selecting various partial decomposition processes or auxiliary decomposition subroutines, there is nothing published on comparing the different decomposition strategies. By general search strategies we understand programmed methods, like for instance deciding whether to execute a decomposition (and which type of), or to continue looking for a better bound set. Using our decomposer MULTIS we found that the general search strategies of the decomposer influence its cost/speed tradeoff more than any other of its single components, such as the encoding algorithm or the column minimization algorithm. In this