Electromechanical couplings play an important role in our day to day lives, as they are present in both natural biological systems and technological applications. Piezoelectricity, the two way coupling between strain and polarization, is one of the most well-known and used mechanism behind electromechanical transduction. This effect is only present in a limited number of materials which are non-centrosymmetric in their atomic structure. Although widely used, piezoelectric materials have important limitations such as their limited operational temperature or their brittleness. In addition, the best and most widely used piezoelectrics have large contents of lead and are thus bioincompatible and unsuited for many biomedical applications due to its toxicity. There exists another electromechanical mechanism far less studied and understood, flexoelectricity, which couples the strain gradient to the polarization, and conversely polarization gradient to strain. Despite piezoelectricity, flexoelectricity is not restricted by symmetry requirements and is therefore a universal property of all dielectrics. Flexoelectricity is noticeable only at small scales where sufficiently large strain and polarization gradients are attainable. Flexoelectricity thus broadens the class of available materials for electromechanical transduction applications and overcomes the limitations of piezoelectric materials, provided that sufficiently small fabrication sizes are achieved. Nevertheless, flexoelectricity has not yet been widely exploited in technology. Towards this goal, new promising piezoelectric metamaterials have been recently proposed which achieve significant apparent piezoelectricity by mobilizing and upscaling the flexoelectric response of small scale non-piezoelectric constituents. The design premise is that the atomic non-centrosymmetry required for apparent piezoelectricity can be replaced by geometrical polarization, i.e. by a noncentrosymmetric material architecture. The performance of these geometrically polarized metamaterials designs is shown to compete in some situations with that of the best known and widely used piezoelectrics and it is expected that it can be further improved through shape optimization. Here, we perform a systematic shape optimization analysis of some recently proposed piezoelectric like metamaterials based on flexoelectricity. Geometric features, lattice orientation and material parameters are optimized in order to maximize four different measures of the apparent piezoelectric response. A rational study of the optimized structure is performed in order to (1) extract the essential design concepts, (2) simplify the geometrical structures and (3) ease future manufacturing. The performance of the optimized designs is compared against two well-known piezoelectrics.