Euclid and LSST are two independent projects which complement each other very well scientifically, but which have had significant problems in working together due to non-scientific reasons. I will first give a head-to-head comparison of the two projects, then discuss several areas where collaboration would significantly improve the results compared to each project working independently. These areas include deblending objects, spectral information for/IR colours of variable sources, galaxy morphologies, improving galaxy ellipticity measurements for weak lensing studies, improved photometric redshifts, more precise cluster mass estimates etc. This talk is based largely on Rhodes et al. (2017, ApJS, 233, 21) but I may add more recent material.