Fast radio bursts (FRBs) are bright, millisecond-duration radio pulses whose origins are unknown. To date, only one (FRB 121102) out of several dozen has been seen to repeat, though the extent to which it is exceptional remains unclear. We discuss detecting repetition from FRBs, which will be very important for understanding their physical origin, and which also allows for host galaxy localisation. We show how the combination of instrument sensitivity, beamshapes, and individual FRB luminosity functions affect the detection of sources whose repetition is not necessarily described by a homogeneous Poisson process. We demonstrate that the Canadian Hydrogen Intensity Mapping Experiment (CHIME) could detect many new repeating FRBs for which host galaxies could be subsequently localised using other interferometers, but it will not be an ideal instrument for monitoring FRB 121102. If the luminosity distributions of repeating FRBs are given by power-laws with significantly more dim than bright bursts, CHIMEs repetition discoveries could preferentially come not from its own discoveries, but from sources first detected with lower-sensitivity instruments like the Australian Square Kilometer Array Pathfinder (ASKAP) in flys eye mode. We then discuss observing strategies for upcoming surveys, and advocate following up sources at approximately regular intervalsand with telescopes of higher sensitivity, when possible. Finally, we discuss doing pulsar-like periodicity searching on FRB follow-up data, based on the idea that while most pulses are undetectable, folding on an underlying rotation period could reveal the hidden signal.