The word can be taken to mean 'fake', but that is a poor interpretation of a more interesting phenomena.
What makes placebo hard to make use of, medically, is both ethical difficulty in delivering it and the variability in effects.
Placebos are often more effective if people are convinced the treatment is good (although there are some exceptions); and this brings up problems of honesty with patients.
The effect of placebos also relates to the nature of the experience, including how intense the experience is: pills and words are typically less powerful than surgery.
Individual people too have varying susceptibility to placebo, which has been linked to genetic variations in neuro-signalling proteins (i.e. COMT).
The efficacy of antidepressants has been put into doubt in the last few years by more careful analysis that properly accounts for placebo effects. Many studies done on antidepressants had missed the fact that people can tell if they are on an active drug, which itself creates an enhanced placebo effect. In research jargon the is also called unblinding. Anything which provides reinforcement could probably enhance placebo effects.
However it is clear that placebo has greater sway over certain realms, particularly those related to inflammation, and to the Brind (brain/mind) such as pain and mood. In contrast it is of little utility for resetting bones (although might help with pain and function!).
Placebo is not simply a case of 'telling stories'. For instance, people who have taken immunosupressant drugs, if given an inert substance in the same situation as the drug was previously given, will have reduced white blood cell counts.
A superficial understanding of placebo has troubled medical research and practice for a long, long time. The imperative to get it right, along with the fascination of it, make placebo well worth investigating further.
Papers:
