Research efforts in zeolite catalysis have become increasingly cognizant of the diversity in structure and function resulting from the distribution of framework aluminum atoms, through emerging reports of catalytic phenomena that fall outside those recognizable as the shape-selective ones emblematic of its earlier history. Molecular-level descriptions of how active-site distributions affect catalysis are an aspirational goal articulated frequently in experimental and theoretical research, yet they are limited by imprecise knowledge of the structure and behavior of the zeolite materials under interrogation. In experimental research, higher precision can result from more reliable control of structure during synthesis and from more robust and quantitative structural and kinetic characterization probes. In theoretical research, construction of models with specific aluminum locations and distributions seldom capture the heterogeneity inherent to the materials studied by experiment. In this Perspective, we discuss research findings that appropriately frame the challenges in developing more predictive synthesis-structure-function relations for zeolites, highlighting studies on ZSM-5 zeolites that are among the most structurally complex molecular sieve frameworks and the most widely studied because of their versatility in commercial applications. We discuss research directions to address these challenges and forge stronger connections between zeolite structure, composition, and active sites to catalytic function. Such connections promise to aid in bridging the findings of theoretical and experimental catalysis research, and transforming zeolite active site design from an empirical endeavor into a more predictable science founded on validated models.