The standard LCDM model of cosmology fares well at explaining large scale structure. However, there are hints that structure has formed more efficiently than is possible in this model. A particular example I will focus on is the collision speed of the components of the Bullet Cluster, though there are a few other examples (e.g. El Gordo). Several studies have shown that it is difficult to reconcile observations of the shock in the gas and the famous separation of dark matter and baryons without a collision speed of ~3000 km/s, mostly in the plane of the sky. Such a high speed appears difficult to reconcile with LCDM. However, the speed is not measured directly. Here, I will describe a method for doing just that. It relies on measuring precisely the redshift difference between multiple images of an object doubly imaged by the Bullet Cluster. The time-dependent gravitational potential of a moving cluster causes this redshift difference (the Moving Cluster Effect). The required accuracy of ~0.1 km/s should soon become feasible with the new generation of radio telescopes, as spatially resolved velocity measurements are not required. I will describe a systematic error that can arise if the source is a rotating spiral galaxy. This is due to uneven magnification across it. However, this can be overcome by looking at the profiles of spectral lines carefully. Thus, I hope to give observers an idea of what to look for in the spectra to be sure they are seeing the Moving Cluster Effect.