Targeted protein degradation is a promising therapeutic strategy capable of overcoming the limitations of traditional occupancy-based inhibitors. By ablating all of the associated functions of a protein at once, the event-driven pharmacology of degrader technologies has recently enabled the targeting of proteins that have been historically deemed "undruggable". Most degradation strategies utilize the ubiquitin-proteasome system to mediate intracellular target degradation and are thus limited to targeting proteins with cytoplasmic domains. While some of these strategies, such as PROTACs, have shown great promise, there is a need for new modalities that can be applied to specifically target cell surface proteins. We previously described the development of an antibody-based PROTAC (AbTAC) that utilizes genetically encoded IgG bispecific antibody scaffolds to bring the cell surface E3-ligase RNF43 into the proximity of a membrane protein of interest (POI) to mediate its degradation. Here, we employ rational protein engineering strategies to interrogate and optimize the properties necessary for efficient degradation of two therapeutically important membrane proteins, PD-L1 and EGFR. We develop multiple antibodies to RNF43 and show that the specific antibody binding epitopes on RNF43 and the POI are more important than the affinities of the AbTAC antibodies. We further expand the available repertoire of E3 ligases by co-opting the E3-ligase ZNRF3 to degrade both PD-L1 and EGFR and show similar importance of epitope for degradation efficiency. Importantly, we show that both RNF43 and ZNRF3 AbTACs do not potentiate unwanted WNT signaling. Lastly, we find that these AbTACs can be even further improved by exploring various dual-binding and IgG scaffolds that range in flexibility, valency, and orientation of the binding arms. These structure-activity and mechanistic studies provide a roadmap for optimizing the development of AbTACs, thereby greatly expanding their utility for targeted cell surface protein degradation.