A drug target is a specific molecule in the body — almost always a protein — whose behavior a medicine is designed to change. In cancer, the target is usually a protein that helps the tumor cell grow, survive, divide, or evade the immune system. The medicine does not act on the cancer as a whole; it acts on one molecule, and the rest of the effect follows from what that molecule does.
A useful way to picture it: a cancer cell is a crowded city, and a target is one specific door that the drug is built to fit. The drug is a key shaped for that door. If the door is on a building the tumor depends on, turning the key can slow the tumor. If the door is also used by healthy tissue, turning it causes side effects.
This is why the choice of target is made before any drug molecule exists. The target defines what the drug must do, what it must avoid doing, and what evidence would count as success. Two properties matter most. First, the target should be genuinely important to the tumor — often the tumor carries an altered or overactive version of the protein, so the tumor depends on it more than healthy cells do. Second, the target should be reachable and changeable by a drug: a protein buried deep inside a cell, or one whose shape offers no place for a drug to bind, is a poor target even if it matters biologically.
A target is therefore not simply 'something cancer needs'. It is a molecule that is important to the tumor, distinct enough from healthy tissue to be tolerable, and physically accessible to a medicine. Those three conditions rarely hold at once, which is why so few candidates survive the selection step.