We need to be ready to measure everything we need for analyses in data using standard signals like Z®mmg and c_c®gJ/y®gmm for photons and Z®ee, J/y®ee and  U®ee for electrons.

Main difficulties here are to understand possible biases, and to estimate uncertainties of extrapolation of measured efficiencies/scales to events with different topology / kinematics.

Here's a link to a study by Y. Gershtein on Z®mmg