This work studies in detail how the judicial application of compliance in parallel manipulators can produce manipulators that require significantly lower actuator effort within a range of desired operating conditions. We propose a framework that uses the Jacobian matrices of redundant parallel manipulators to consider the influence of compliance both in parallel with the actuated joints as well as the passive joints, greatly simplifying previous approaches. We also propose a simple optimization procedure to maximize the motor force reduction for desired regions of the workspace and range of external forces. We then apply the method to a Stewart-Gough platform and to a 3-URS (universal rotational and spherical joint) manipulator. Our results show that parallel manipulators with tasks that involve a preferred external force direction, as for instance, big weights in the platform, can see large reductions in actuator effort through the judicial use of compliant joints without significantly losing rigidity.