We present an in-depth analysis of the experimental estimation of cross-sections in surface enhanced raman scattering (SERS) by vibrational pumping. The paper highlights the advantages and disadvantages of the technique, pinpoints the main aspects and limitations, and provides the underlying physical concepts to interpret the experimental results. Examples for several commonly used SERS probes are given, and a discussion on future possible developments is also presented. Obtaining good estimates of SERS cross-sections is, in general, an extremely hard problem and has been a longstanding ambition of the SERS community for reasons that go from the purely applied (quantification of signals) to the more fundamental (comparisons of theoretical electromagnetic enhancement factors with experiment). Any method that can produce a standard protocol for the estimation of cross-sections is, accordingly, of great interest and an effort to understand its principles and limitations is required.