Traditional operational transconductance amplifier (OTA) optimizers decouple sizing from layout synthesis, leading to poor results under layout-induced parasitics—a problem exacerbated in FinFET nodes. This work proposes a hierarchical methodology for FinFET nodes that maintains self-consistency between schematic-level and postlayout design, leveraging a SPICE-calibrated gm/ID methodology that incorporates layout parasitics. Unlike machine-learning-based approaches that rely on heavy SPICE simulations and large datasets, our approach is computationally efficient and invokes SPICE sparingly. Experimental results across four OTA topologies demonstrate that performance of the sized circuit is greatly degraded after layout (up to 44% in UGF, 17 dB in DC gain, and 22 dB in CMRR), yet the systematic, closed-loop use of correction factors with self-consistent gm/ID methodology enable recovery of postlayout performances and yields designs that meet specifications. The full flow requires only 10-15 full SPICE simulations, enabling thorough exploration of design space for high-quality solutions.