Flexible and highly sensitive piezoresistive nanocomposites have been demonstrated to possess considerable potential for monitoring structural integrity and human physiological performance. To enhance the mechanical and strain sensing properties of these nanocomposites, different nanofillers (e.g., metal nanowires, carbon nanotube, and graphene) have been incorporated in polymeric matrices to establish electrically conductive pathways that are also sensitive to applied strains. Their piezoresistivity mainly stem from nanofillers’ intrinsic piezoresistivity, tunneling effect, and contact resistance changes of the nanofiller networks. Although many high-performance nanocomposite strain sensors have been developed and using different techniques, the empirically guided fabrication approach can be laborious, inefficient, and, most importantly, unpredictable. Therefore, this study proposes a topological design-based approach to strategically control and manipulate the strain sensing performance of the nanocomposites, simply by altering its geometric pattern design. First, polyethylene terephthalate (PET) substrates were patterned with pre-designed hierarchical inhomogeneous topologies and kirigami cuts created using a laser cutter. Second, the substrates were spray-coated using a carbon nanotube (CNT)-latex to deposit the strain-sensitive thin films. Third, the strain sensing performance of the CNT-latex nanocomposite thin films of different topologies was characterized and compared. It was found that, as the initial solid mechanics analysis predicted, the hierarchical inhomogeneous topology effectively enhanced the nanocomposites’ strain sensitivity, while the kirigami cuts significantly reduced sensitivity. The proposed methodology can help guide the development of high-performance nanocomposites with pre-programmed sensing properties for structural and human health monitoring applications.