A basic principle of photometry goes back to Steinheil, who built the world's first stellar photometer in the early 1830's. Steinheil's principle is that only similar things should be compared. That is, good results are obtained when as many variables as possible are kept fixed. This minimizes the number of instrumental parameters that have to be calibrated, and allows the necessary calibrations to be a small part of the total data gathered. All observations must be made with the same instrument, used under the same conditions.
However, we cannot make all our observations at the same time, or at the same place in the sky. So we must be able to separate different effects that can vary with time, such as extinction coefficients and instrumental zero points. To do this, we must distribute the observations so that the main independent variables (such as airmass, star color, and instrumental temperature) are uncorrelated with each other. In particular, they should all be uncorrelated with time, so far as possible. As some (like temperature) are inherently likely to be correlated with time, it is doubly important that others (like airmass) be uncorrelated. Achieving this condition requires a certain amount of advance planning.