I will present two scenarios where standard searches for SUSY at the LHC are not very efficient because the mass splitting between the LSP and the NLSP is small and therefore there is not a lot of MET in the events. I will then use different observables to be able to improve the reach of the LHC in this kind of situations.