The article explores the implications for public welfare policy of the current reliance on quantification tools to assess both people and programs. In this context, profiling has recently acquired a key role in active labour market policies, because it allows policy administrators to distinguish among different levels of risk in individuals approaching the labour market. After discussing the issue of the deployment of numbers in policies, I present and contextualize a recent case study: a national profiling system adopted in Italy in fulfilment of Youth Guarantee, a European public policy whose aim is to support young people entering the labour market via dedicated services. In the second part of the article, I outline how the profiling system was devised and how it works, going on to report a qualitative study on the implications of using the Disadvantage Index - the core of the profiling model – and how it was modified after an initial test phase. The research findings suggest that, in contradiction with the Index’ original goals, it has ended up screening out the more disadvantaged NEETs, with the risk of perpetuating their disadvantage. In the conclusions section, I return to the role of numbers and indicators in governing public policy, as well as their contribution to the depoliticization of public action.