Chiral shift reagents are a real pain.
There really isn't a trick to it. Take an NMR of some racemic material, then start adding your shift reagent in small portions and take an NMR after each addition. Look for the diastereomeric complexes to start separating. If it works, then you will want to try to find out what sort of concentration of shift reagent you need to observe adequate separation in the NMR so that you can repeat the assay on other samples. If you add too much shift reagent, you can easily overshoot the optimal separation and get other peaks to start overlapping, so be careful.