The detection of a neutron star merger by the Advanced Laser Interferometer Gravitational-Wave Observatory and Advanced Virgo gravitational wave detectors, and the subsequent detection of an electromagnetic counterpart have opened a new era of transient astronomy. With upgrades to the Advanced Laser Interferometer Gravitational-Wave Observatory and Advanced Virgo detectors and new detectors coming online in Japan and India, neutron star mergers will be detected at a higher rate in the future, starting with the O3 observing run which will begin in early 2019. The detection of electromagnetic emission from these mergers provides vital information about merger parameters and allows independent measurement of the Hubble constant. The Australian Square Kilometre Array Pathfinder is expected to become fully operational in early 2019, and its 30 deg2 field of view will enable us to rapidly survey large areas of sky. In this work we explore prospects for detecting both prompt and long-term radio emission from neutron star mergers with Australian Square Kilometre Array Pathfinder and determine an observing strategy that optimises the use of telescope time. We investigate different strategies to tile the sky with telescope pointings in order to detect radio counterparts with limited observing time, using 475 simulated gravitational wave events. Our results show a significant improvement in observing efficiency when compared with a naïve strategy of covering the entire localisation above some confidence threshold, even when achieving the same total probability covered.