There has been considerable debate surrounding the evolution of eusociality, which has recently increased in vigor with regard to what actually constitutes eusociality. Surprisingly, there has been little discussion on terminologies for describing social systems that are more-or-less egalitarian, yet such societies form an obvious contrast to eusociality, and transitions between these two forms of social organization appear to be common. We argue that current terminologies and methods for dealing with non-hierarchical societies are not well suited for such comparative approaches to social evolution. We outline three problems for comparative approaches (identifying egalitarianism, implied egalitarianism and taxon-specific terminology) and propose two solutions. The first solution is a re-sampling method to assess investment asymmetries, and the second is the introduction of the term “casteless” to encompass forms of social organization where there is no lifetime commitment to queen-like or worker-like roles, but where skew in reproduction or alloparental tasks may nevertheless be apparent at any one time. Our suggested terminology avoids the implied egalitarian nature behind the terms communal and quasisocial, which place undue emphasis on specific nesting biologies and which have the potential to impede ‘bottom-up’ comparative studies of social evolution. Such non-eusocial groups provide the best insights for understanding how social behavior evolved and our suggested approaches should enhance future investigations.