Our Milky Way provides a unique test case for galaxy evolution models because of our privileged position within the Milky Way's disc. This position also complicates comparisons between the Milky Way and external galaxies, due to our inability to observe the Milky Way from an external point of view. Milky Way analogue galaxies offer us a chance to bridge this divide by providing the external perspective that we otherwise lack. However, overprecise definitions of 'analogue' yield little-to-no galaxies, so it is vital to understand which selection criteria produce the most meaningful analogue samples. To address this, we compare the properties of complementary samples of Milky Way analogues selected using different criteria. We find the Milky Way to be within 1{sigma} of its analogues in terms of star formation rate and bulge-to-total ratio in most cases, but we find larger offsets between the Milky Way and its analogues in terms of disc scale length; this suggests that scale length must be included in analogue selections in addition to other criteria if the most accurate analogues are to be selected. We also apply our methodology to the neighbouring Andromeda galaxy. We find analogues selected on the basis of strong morphological features to display much higher star formation rates than Andromeda, and we also find analogues selected on Andromeda's star formation rate to overpredict Andromeda's bulge extent. This suggests both structure and star formation rate should be considered when selecting the most stringent Andromeda analogues.