The article discusses the challenges in resolving difficult phylogenetic questions using phylogenomics, a field that relies on genome-scale data to infer evolutionary relationships. Despite the increasing availability of sequencing technology, the authors argue that simply adding more sequences is not enough to resolve inconsistencies in phylogenetic trees. They highlight three recent large-scale analyses of early animal diversification, which produced highly incongruent findings despite extensive sequence data. The authors identify two main factors contributing to the difficulty in reconstructing accurate phylogenetic trees: (1) closely spaced speciation events leading to short internal branches, and (2) ancient events with multiple substitutions, resulting in homoplasy. They emphasize the importance of reducing non-phylogenetic signal, which can arise from incorrect identification of orthologous genes, erroneous alignments, and misinterpretation of multiple substitutions. The article provides strategies to mitigate these issues, including improving alignment quality, using realistic models of sequence evolution, and increasing taxon sampling. The authors conclude that while phylogenomics can be robust to the inclusion of non-orthologous sequences when genuine phylogenetic signals are abundant, it is sensitive to non-phylogenetic signals, especially when phylogenetic signals are weak. They advocate for a comprehensive approach that addresses all sources of non-phylogenetic signal to improve the accuracy of phylogenetic inference.The article discusses the challenges in resolving difficult phylogenetic questions using phylogenomics, a field that relies on genome-scale data to infer evolutionary relationships. Despite the increasing availability of sequencing technology, the authors argue that simply adding more sequences is not enough to resolve inconsistencies in phylogenetic trees. They highlight three recent large-scale analyses of early animal diversification, which produced highly incongruent findings despite extensive sequence data. The authors identify two main factors contributing to the difficulty in reconstructing accurate phylogenetic trees: (1) closely spaced speciation events leading to short internal branches, and (2) ancient events with multiple substitutions, resulting in homoplasy. They emphasize the importance of reducing non-phylogenetic signal, which can arise from incorrect identification of orthologous genes, erroneous alignments, and misinterpretation of multiple substitutions. The article provides strategies to mitigate these issues, including improving alignment quality, using realistic models of sequence evolution, and increasing taxon sampling. The authors conclude that while phylogenomics can be robust to the inclusion of non-orthologous sequences when genuine phylogenetic signals are abundant, it is sensitive to non-phylogenetic signals, especially when phylogenetic signals are weak. They advocate for a comprehensive approach that addresses all sources of non-phylogenetic signal to improve the accuracy of phylogenetic inference.