Clicker training is based on the process of "Operant Conditioning" (according to Skinner ) and is prepared by "Classical Conditioning" according to Pavlov.
The clicker training uses the effect that a behavior is shown all the more often, if on the behavior a consequence takes place, which represents a (colloquially expressed) "profit" for the animal (learning by positive reinforcement in the form of reward, and/or "learning at success" after Thorndike).
A "gain" for an animal is e.g. food or play.
Essential for "learning by success" is that the animal experiences the consequence of its behavior within as short a time as possible, so that the connection between behavior and consequence (the linkage) is maintained (the linkage time window for e.g. dogs is in the range of maximum 1-3 seconds). The click itself is not a reward, but only a marking signal for the precise moment of the desired behavior, like a photographic snapshot ... a promise for the coming reward.