It is actually more of a structural comparison. The idea is to mimic the interaction between that functional group and the specific site on the protein with which it is associated.
For example, if you have a molecule that contains a carboxylic acid, it might be interacting with the amine group on a specific lysine residue in the protein. Replacing it with another group that maintains the flat geometry of the carboxylic acid and the pH of the carboxylic acid might allow you to maintain the same interaction between that point on the molecule and the protein. X-ray crystallography can be used to confirm the same binding characteristics. Tetrazoles are basically flat in the same way carboxylic acids are, have a hydrogen with approximately the same pH, and have lone pairs of electrons oriented in similar directions to the lone pairs on the oxygens of a carboxylic acid. If there are no steric constraints that prevent the tetrazole from moving into the same position the carboxylic acid would occupy, the tetrazole can function as a bioisostere - the molecule would bind to the enzyme in the same position and with the same strength.