Depression is one of the most burdensome mental disorders, and effective treatments for depression are needed. This review highlights that the network perspectives of psychopathology can promote tailored therapies for depression and improve the quality of treatment. First, we discuss that depression heterogeneity can be depicted by quantitative data analyses. Second, we review two previous clustering approaches that have been utilized to develop tailored therapy for depression: the first is based on a staging model and the second focuses on depression subtypes. Third, we focus on models and analyses of psychopathology networks, which have recently received substantial attention in clinical psychology. We demonstrate that psychopathology network models view mental disorders as complex interplays of symptoms, and we introduce analytic procedures and previous studies based on these models. We then summarize differences and similarities between network and clustering approaches and discuss how psychopathology networks can further promote tailored therapies for depression. Finally, we discuss the requirements for the practical use of psychopathology networks from the perspectives of cost-effectiveness and collaboration of data scientists and clinicians.