Social deficits are a defining characteristic of youth with autism spectrum disorder (ASD); however, symptom presentation is often highly heterogeneous (Georgiades et al., 2013). While research on interventions to remediate such deficits among individuals with ASD has recently proliferated, results remain similarly variable (Gates et al., 2017). This variability begs the fundamental questions of “for whom” do current treatments work (Kazdin, 2007). Likewise, in parallel to effort across child psychopathology (Ng & Weisz, 2017), there have been calls to individualize interventions in ASD in order to maximize benefits for participants (Lerner et al., 2012). The question remains, though: where – among the vast array of potential factors – to begin? In fact, basic psychosocial and demographic information is available in nearly all intervention datasets that contains under-explored factors that may be used to constrain heterogeneity of outcomes and potentially individualize interventions. This panel will synthesize findings from several social-focused interventions to explore when and how widely-available key moderators – age, gender, language, and comorbid diagnoses – may similarly and differentially impact outcomes. An integrative discussion will address both person- and treatment-level factors that may drive variability in moderator effects and provide concrete direction for the future of ASD social interventions.