Quantitative flow velocimetry in Optical Coherence Tomography is used to determine both the axial and lateral flow component at the level of individual voxels. The lateral flow is determined by analyzing the statistical properties of reflected electro-magnetic fields for repeated measurements at (nearly) the same location. The precision or statistical fluctuation of the quantitative velocity estimation depends on the number of repeated measurements and the method to determine quantitative flow velocity. In this paper, both a method to determine quantitative flow velocity and a model for the prediction of the statistical fluctuations of velocity estimations are developed to analyze and optimize the estimation precision for phase-based velocimetry methods. The method and model are validated by phantom measurements in a bulk scattering medium as well as in intralipid solution in a capillary. Based on the model, the number of repeated measurements to achieve a certain velocimetry precision is predicted.