Drug discovery is not one activity but a sequence of stages, each with its own goal and its own way of failing. It begins with target selection: researchers decide which biological molecule in the body, usually a protein, they want to influence. A protein is a large folded molecule that carries out specific jobs in cells, and a drug target is the particular protein whose activity the future medicine is meant to change.
Once a target is chosen, the search for a hit begins. A hit is any molecule that shows some measurable effect on the target. Hits are then refined into leads, molecules with better potency (the strength of the effect) and better drug-like behavior. This refinement stage is called lead optimization.
Only then does the compound enter preclinical testing, where it is examined in cells and animals for safety and for how the body absorbs and clears it. If it survives, it moves into human testing, which runs in three phases: phase 1 checks safety in a small group, phase 2 looks for signs of effectiveness, and phase 3 compares the drug against existing treatment in a large population. If the results hold, the developer submits the compound for regulatory approval, and only after approval does it reach patients.
The important structural fact is that these stages are sequential and gated. A compound that fails at any gate is discarded, and the money and years already spent on it are not recovered.