Gravitational lensing has long been considered as a valuable tool to determine the total mass of galaxy clusters. The shear profile as inferred from the statistics of ellipticity of background galaxies allows to probe the cluster intermediate and outer regions thus determining the virial mass estimate. However, the mass sheet degeneracy and the need for a large number of background galaxies motivate the search for alternative tracers which can break the degeneracy among model parameters and hence improve the accuracy of the mass estimate. Lensing flexion, i.e. the third derivative of the lensing potential, has been suggested as a good answer to the above quest since it probes the details of the mass profile. We investigate here whether this is indeed the case considering jointly using weak lensing, magnification and flexion. We use a Fisher matrix analysis to forecast the relative improvement in the mass accuracy for different assumptions on the shear and flexion signal - to - noise (S/N) ratio also varying the cluster mass, redshift, and ellipticity. It turns out that the error on the cluster mass may be reduced up to a factor 2 for reasonable values of the flexion S/N ratio. As a general result, we get that the improvement in mass accuracy is larger for more flattened haloes, but extracting general trends is a difficult because of the many parameters at play. We nevertheless find that flexion is as efficient as magnification to increase the accuracy in both mass and concentration determination.