Somewhere in almost every article about hypnosis, including ours, there’s a sentence like this one: about 10 to 15% of adults respond dramatically, around 70% respond moderately, and about 15% respond very little. It gets repeated because it’s useful and because it’s roughly true. It rarely gets sourced, and it almost never gets explained.
Those numbers are score bands on a test. They come from a measurement tradition that is older than most of the therapies it now gets applied to, and understanding where they come from changes what you do with them. If hypnosis has done nothing for you, the interesting question isn’t whether you were trying hard enough. It’s whether you’re being asked to do something your nervous system does easily or something it doesn’t do much at all.
What the trait actually measures
Hypnotic suggestibility, also called hypnotisability, is how strongly a person responds to suggestion after a hypnotic induction. That’s the entire definition, and the narrowness is deliberate. It doesn’t mean you believe things easily. It doesn’t mean you can be talked into a bad decision. It doesn’t mean you go blank or lose control.
What it means is specific and slightly strange: when a voice says your arm is becoming heavy, does your arm become heavy? Not “can you pretend it does,” and not “do you feel relaxed.” Does the described experience arrive, without you assembling it on purpose. Highly hypnotisable people report that it simply happens to them. Low responders report understanding the instruction perfectly and nothing following from it.
This is why the trait is so poorly served by the word suggestible. In everyday English that word means gullible, and the research has consistently failed to find that link. Interrogative suggestibility, the tendency to change your story under leading questions, is measured separately and correlates weakly with hypnotic response. Compliance and conformity don’t predict it either. The one thing that does, modestly, is absorption: the ordinary capacity to become so involved in a novel, a film, or a landscape that the room drops away. Auke Tellegen and Gilbert Atkinson described that trait in 1974 and it has been the most reliable personality correlate of hypnotisability ever since, though the relationship is loose enough that you can’t predict one from the other in any individual case.
Where the responder numbers come from
Two scales did almost all of this work, both built in the same few years by the two research groups that took hypnosis seriously when very few people did.
The Stanford scales
André Weitzenhoffer and Ernest Hilgard published the Stanford Hypnotic Susceptibility Scale in 1959, with Forms A and B, and the harder Form C in 1962. The design is unglamorous and that’s the point. A standardised induction is read out, followed by a fixed list of twelve suggestions, and an observer records whether each one produced a response according to a stated criterion. Your hand moves apart from the other, or it doesn’t. Your eyes stay closed when you’re told you can’t open them, or they open. You score from 0 to 12.
The items climb in difficulty. Early ones are motor and easy: postural sway, eye closure, an arm growing heavy and lowering. Middle ones involve inhibition, like being unable to unlock your interlocked fingers. The hardest ones ask for hallucination and amnesia, hearing a voice that isn’t there, or being briefly unable to recall what happened until a cue releases it. Almost everyone passes some of the early items. Very few people pass the last ones.
The Harvard group version
The Stanford scales require one experimenter per person, which makes population data expensive. Ronald Shor and Emily Orne solved that in 1962 with the Harvard Group Scale of Hypnotic Susceptibility, Form A, which runs from a recording to a room full of people at once and has each of them score their own responses afterwards. It is the workhorse. Most of the large samples in the literature, and therefore most of the distributions quoted in articles like this one, are Harvard Group Scale scores.
Plot enough of those scores and you get a broad, slightly lopsided spread across the full range with a bulge in the middle. Split it into low, medium and high bands and you land somewhere near the familiar figures. So the 10–15–70–15 shape isn’t a finding about human nature so much as a summary of where people fall on a twelve-item test. That’s a weaker claim than it’s usually made to sound, and it’s still the best guide available.
That clip is a loose version of one of the easier scale items. A real administration is scored against a stated threshold, in this case whether the arm lowers a specified distance within a set time, and nothing about it is atmospheric. The lack of theatre is what makes the numbers mean anything.
It barely moves over a lifetime
The finding that ought to change how you plan is the stability one. In 1989 Carlo Piccione, Ernest Hilgard and Philip Zimbardo tracked down Stanford participants who had been scored 25 years earlier and tested them again. The correlation between scores across that quarter century was around 0.71, which puts hypnotisability among the most stable individual differences psychology has measured, closer to cognitive ability than to anything mood-shaped or attitudinal.
The one systematic movement is developmental. Measured suggestibility rises through childhood, appears to peak somewhere around ages nine to twelve, and settles lower in adolescence. After that it stays where it lands.
People do ask whether it can be trained, and the answer is a genuinely contested “a little, perhaps.” The Carleton Skill Training Program, developed by Nicholas Spanos and colleagues in the 1980s, reported that low responders could be moved substantially upward by teaching them how to interpret and engage with suggestions. Critics argued the gains reflected participants learning what a compliant subject looks like rather than a change in underlying response, and that argument was never fully settled. Attentional training and expectancy manipulation both produce small effects. None of it is reliable enough to build a nightly practice on. Treat the trait as close to fixed, work around it, and you’ll be right more often than not.
If you score low, that’s a routing instruction
Here is the part the cluster owes you, and the reason this page exists. A low score is not a failure of effort or belief. It is a fact about how your attention behaves, it was probably true when you were nineteen, and it will likely be true at sixty. What follows from it is practical.
Start by separating what a hypnosis recording actually contains. A sleep hypnosis track is at least four things stacked together: slow speech that entrains your breathing, a progressive relaxation pass through the body, a voice that occupies the channel your own rumination would otherwise use, and a layer of suggestion. Only the last of those depends on suggestibility. Low responders who report that hypnosis tracks “do nothing” often mean the suggestions don’t land, while quietly getting real benefit from the other three. Before you abandon the format, check which part you were judging.
If it genuinely isn’t working, the cluster splits cleanly along this line. Practices that require a suggestive response include hypnosis in all its forms and, to a lesser extent, self-hypnosis for sleep and clinical hypnotherapy, where the correlation between hypnotisability and outcome is strongest for hypnotic pain control and looser elsewhere. Practices that don’t require it work by repetition, by attention, or by physiology: sleep affirmations, which get their effect from familiarity rather than from trance; body scan meditation and progressive muscle relaxation, which are somatic and ask nothing of your imagination; 4-7-8 breathing and cognitive shuffling, which are mechanical enough to work whether or not you’re on board with them.
Notice what’s absent from that second clip. No descent, no imagery, no described sensation you’re expected to have. It’s a sentence and a schedule. That’s the shape of a practice built for someone the scales would put in the bottom band, and it’s the shape most of Murmora’s own audio takes.
The mirror-image point is worth making too, because high responders get bad advice from the other direction. If suggestion lands hard for you, you’re the group for whom expectation and framing do the most work, and you should be correspondingly careful about what you agree to have suggested. A track that repeats a claim you don’t actually want to hold is not neutral for you the way it might be for someone else.
The smallest version of the test
You don’t need a scored scale. You need two weeks and one honest criterion set in advance.
Pick one hypnosis recording. Not a new one each night, because novelty confounds the result. Same track, same time, fourteen nights. Before you start, write down the single thing you’d expect to change if this worked for you, and make it observable: falling asleep before the track ends, or noticing the described heaviness in your hands, or the specific 3 a.m. thought loop not starting. “Feeling better” is unfalsifiable and will not tell you anything.
At the end of two weeks, sort yourself into three outcomes. Something clearly shifted, in which case keep going and the trait question is answered well enough. Nothing at all happened, in which case take the routing above seriously and move to a practice that doesn’t run on suggestion. Or the relaxation helped and the suggestions felt like words in the dark, which is the most common result and points at the middle of the distribution. For that third group the fix is usually content rather than technique: generic suggestions land poorly on moderate responders, and specific ones about your actual situation land better. That’s the same calibration problem how to write affirmations works through in detail.
How Murmora fits
Most of what Murmora makes is deliberately not hypnosis. The sparse-whisper format is short, plain sentences with long silences between them, repeated on a schedule, which is a design that leans on familiarity rather than on trance. If you’re in the bottom band of the distribution, that matters: it’s a format that doesn’t require you to respond to a suggestion for anything to accumulate.
The other half of the answer is specificity. Moderate responders, which is most people reading this, are the group where generic content is the binding constraint rather than the trait. Murmora builds the lines from what you describe about your own situation and speaks them in one of a few guide voices, or later in a clone of your own. If two weeks of a stock track told you the mechanism engages for you but the words don’t, that’s the variable worth changing next.