Retrosynthetic analysis can be compared to the climbing of a tree. You are at the root of the tree (the target molecule) and want to climb to the top branches (the starting materials). The tree branches off many times. You have to choose your way! Which way is the fastest, the easiest and the most reliable to the best starting materials? E.J. Corey has developed and analysed many different strategies for making such a choice. One might e.g. choose to use a specific reaction. Alternatively, one might choose to build (synthesize) a particular part of the target molecule first or choose to solve some specific problem associated with the three-dimensional structure of the target, its stereochemistry.
Today, the analysis can be performed with computers in which an ever increasing number of reactions are stored. The computer gives suggestions, but the chemist must choose the way and reactions from his/her own qualifications and experience.