There are a number of ways of controlling calibration costs. First, by making sure that all the appliances impacting the quality of the measurement results or the quality of the product are properly calibrated, and that only these appliances are calibrated. To this end, it is always useful to draw up a list of the components of uncertainty and to quantify them so that you can demonstrate that a calibration is unnecessary due to the limited impact of the instrument on the final result. Then, by using the certificates of successive calibrations to deduce, through experience, the interval between two calibrations that achieves the best compromise between risk and frequency.