We have the conventional definition of "placebo" versus the Patrick Wall definition, which is the same definition used by PTs interested in treating pain.
For anyone who still doesn't know who Patrick Wall was, he and Ronald Melzack combined efforts to research pain for over 4 decades. (Tip: Read the paper linked to Ronald Melzack. It explains his neuromatrix theory of pain. It's the best pain theory in existence to date, in my humble opinion.)
But now, back to placebo. Placebo is a loaded word, charged with centuries of flimflam, misunderstanding, and exploitation. Patrick Wall figured out what placebo response really is, how that works, and worked together with PTs to determine how it could be harnessed ethically for relief of pain.
I count myself among this slowly growing cadre of PTs who are laying aside our old tissue-based belief system based on 3 or 400 years of ignorance about pain, in favor of instead understanding the science that has developed around pain. Patrick Wall said (paraphrased), "Placebo is not something administered TO a patient, it is something to be elicited FROM a patient." He went on to describe the perfection of a placebo response. It was something the patient's brain made by itself, a chemical substance, antidote, precise in dosage and specific to the problem, which went straight to the receptors in need of it. It was allowed to exist in the brain for only as long as necessary, whereupon the brain would dismantle it by itself. In other words, one sort of nature (placebo response) taking care of another kind of nature (pain response).
After a brief nervous system explanation, it's a lot easier to explain effects that are "placeboic" in a way such that people will be more inclined to think of "placebo" in a positive light rather than negative - at least they catch a glimpse that it will be good for pain, that you don't think it's undesirable, that you want to help them make their very own, and begin to see achieving it as a victory not a defeat. It must be reframed/ redefined as a treatment effect that is desirable and unique to them, something that is produced naturally as a consequence when a "team" (comprised of patient and therapist) develops a temporary third entity (the interaction of nervous systems) to help a fourth entity (the patient's own nervous system) wrestle with and overcome a fifth, the "foe" (pain output). When people have the Wall definition of placebo explained to them they are not at all disinclined to go for it. They are willing to set up a treatment improv mini-drama with you in which all these entities can exist on their own for a short while, and change places, play musical chairs until all the chairs are taken and pain has nowhere to sit. They become co-conspirators in the development of a "placebo" effect that will fix their own system, in fact they get that it is up to them, by doing very little except waiting patiently and tracking processes. The right frame around that charged concept, "placebo", is everything.
The team can be dissolved. The patient can go off with new strength derived from new cognition around the problem, a new understanding that while their nervous system includes "them", it is not "them" entirely. Instead it is something they can successfully interact with, as one might find oneself interacting with an unruly child or a screaming baby. Do not abuse the nervous system. Do not permit anyone else to abuse it. Figure out what it needs (usually some form of movement developed slowly and carefully, with close attention to a sense of timing), then supply it, and be patient. Do this for short periods of time (minutes), frequently (as one would feed a starving baby, perhaps every hour or two). It will take a good three days for abnormal and too numerous receptor sites (associated with pain sensitization) to dismantle and (hopefully better ones) to reform. Allow time for improvement to establish itself. Get on a better track and find ways to enjoy the rest of your hours per day while waiting optimistically.