It is well established that the Marangoni flow dominated circulation within a laser melt pool significantly modifies the pool profile and temperature distribution. Detailed computational fluid dynamics models are required to accurately predict this but these are complicated and computationally expensive. Many researchers have in the past used an enhanced thermal conductivity approach, but the validity of this approach for accurately predicting the melt pool geometry and temperature distribution is largely unproven. This paper presents an analysis of the widely-used isotropic enhanced thermal conductivity approach and compares it with a more advanced anisotropic approach for modelling the laser melting of Inconel 718. Experimental and modelled results for the geometry of a melt pool created by a moving laser beam are compared. It is found that the conventional enhanced thermal conductivity approach does not change the melt pool size and shape; it only reduces the maximum surface temperature. The anisotropic enhanced thermal conductivity approach on the other hand is able to modify the melt pool size and geometry and yields a better agreement with the experimental results.