 The main potential is on integrating them with traditional sources like census data in order to fill temporal gaps and improve the spatial resolution at which migration is usually recorded and this integration would also allow to leverage on strengths of each data and at the same time minimizing some of their bias. We are using cell phone data records and for interview along with machine learning techniques to identify gender specific pattern in a ball in order to support policy. Furthermore, in collaboration with IOM and the Internal Displacement Monitoring Center we are also combining cell phone data records with displacement tracking matrices in order to better understand disaster driving migration. So the main advantage is like their massive sample size, their relative arrival origins and destination and length of stay that we can derive from them and also the very detailed spatial and temporal resolution and the fact that from at least some of them we can retrieve information about social network and wealth. In terms of the limitation we have to be aware that some of the groups may not be represented such as the elderly, the poorest and the children and the fact that they not explicitly contain demographic information. Another limitation is given by the difficult to assess and share the data because of mostly of commercial and privacy concern.