In order to capture the heterogeneity of ideas about family that contributes to shaping family change, we build and analyse a dataset composed of over half a million individuals to develop a new set of indicators based on role identities - as displayed in Twitter profiles. We compute our indicators for US States and show how they relate to behavioural indices linked to the ideas of Second Demographic Transition and Gender Revolution. Our approach provides an innovative way to map, and potentially track over time, the cultural footprints that underlie family change.