Over the last years, the availability of genomic sequence data from thousands of different species has led to hopes that a phylogenetic tree of all life might be achievable. Yet, the most accurate methods for estimating phylogenies are heuristics for NP-hard optimization problems, many of which are too computationally intensive to use on large datasets. Divide-and-conquer approaches have been proposed to address scalability to large datasets that divide the species into subsets, construct trees on subsets, and then merge the trees together. Prior approaches have divided species sets into overlapping subsets and used supertree methods to merge the subset trees, but limitations in supertree methods suggest this kind of divide-and-conquer approach is unlikely to provide scalability to ultra-large datasets. Recently, a new approach has been developed that divides the species dataset into disjoint subsets, computes trees on subsets, and then combines the subset trees using auxiliary information (e.g., a distance matrix). Here, we describe these strategies and their theoretical properties, present open problems, and discuss opportunities for impact in large-scale phylogenetic estimation using these and similar approaches.