It Matches Their Current Experience
Begin with something the listener already knows to be true. “You have been considering this decision for some time.” Agreement earned here costs you nothing and buys you everything that follows.
The NLP Milton Model
People rarely resist change. They resist being told exactly what to think, feel and do. The Milton Model is the language that removes the instruction and leaves the possibility — deliberately open, deliberately permissive, and considerably harder to use well than it looks.
In Short
The Milton Model is a set of language patterns in NLP that deliberately leaves selected information unspecified, so the listener fills the space with their own meaning. It is the mirror image of the Meta Model. Where the Meta Model asks precise questions to recover what a sentence left out, the Milton Model constructs sentences that leave things out on purpose. Compare “Stop worrying and feel confident” with “You may begin to notice that confidence grows when you approach this differently.” The first issues an instruction. The second creates room. Used well, it lowers resistance and invites reflection. Used badly, it is vague wordplay — or manipulation.
Written And Reviewed By
Founder And Lead Trainer, NLP Limited
Rajiv Sharma is an NLP Master Trainer certified through The Society of NLP (USA), trained and coached directly by Dr. Richard Bandler, co-creator of Neuro-Linguistic Programming. He holds the ICF PCC credential and created the MARK Model® and the LeadFORTH framework. He is the author of Make a MARK in Life and AI-ENABLE for Sales. Over a career beginning in 1990, he has trained more than 850,000 professionals across 57 countries. His work is endorsed by Marshall Goldsmith, Brian Tracy and John Mattone.
Why It Works
This is the distinction everything else depends on, and it is the one most people get wrong. Ordinary vagueness produces confusion. Artful vagueness produces room.
Merely Vague
“Something good will happen.”
No direction. No attention. Nothing for the listener to do with it. This is a fortune cookie.
Artfully Vague
“As you think about your next step, you may begin to recognise an opportunity you had not considered before.”
Still open — and yet it points attention at the future, at possibility, at discovery. The listener decides what the opportunity is. The sentence decides that there is one to look for.
Both sentences withhold detail. Only one of them is doing a job. The language is vague; the purpose must not be. A Milton Model sentence built without a clear outcome behind it is not influence — it is word salad with a soothing voice.
The model works because meaning is made by the listener, not delivered by the speaker. When you say “you may discover a better approach”, the listener has to supply an approach from their own experience in order to understand the sentence at all. Whatever they supply is theirs. They will defend it, act on it and remember it in a way they would never do with an approach you handed them. This is the same mechanism described in the NLP Communication Model — running deliberately, and in reverse.
New to the subject? Start with What Is NLP?
The Name
It is named after Dr Milton H. Erickson, an American psychiatrist known for his work in medical hypnosis and psychotherapy. Erickson worked through stories, metaphor, indirect suggestion and permissive language rather than direct instruction. Richard Bandler and John Grinder studied his communication closely and described the recurring patterns in Patterns of the Hypnotic Techniques of Milton H. Erickson, M.D.
An honest caveat. The Milton Model is not a complete account of what Erickson did. It is an NLP interpretation of certain patterns found in his language. Erickson himself worked from relationship, timing and a degree of clinical judgement that no list of twenty-five patterns can contain. Learn the patterns. Hold them lightly.
The pairing matters. The Meta Model was built by modelling therapists who asked ruthlessly precise questions. The Milton Model was built by modelling a man who almost never did. Together they form the two halves of NLP's language work — one moving towards precision, one towards possibility. A practitioner who has only one of them has half a toolkit.
The Mechanism
Strip away the twenty-five pattern names and this is what remains. Every well-formed Milton Model sentence is doing some combination of these five things.
Begin with something the listener already knows to be true. “You have been considering this decision for some time.” Agreement earned here costs you nothing and buys you everything that follows.
Possibility, not command. “You might begin to see this differently.” Nothing is being demanded, so there is nothing to push back against.
“You can discover the approach that works best for you.” Which approach? Unstated — and that is the point. The listener supplies one, and it will fit them better than anything you could have named.
Open is not aimless. “You may notice the resources you already have.” The sentence does not say which resources. It does insist there are some, and sends the listener looking.
“You can consider this and decide whether it is useful for you.” The listener keeps the right to accept, reject or reinterpret. Remove this one and you have stopped doing influence and started doing something else.
The Twin
They are not rivals. They are the two directions language can move, and a capable communicator chooses between them several times in a single conversation.
| The Meta Model | The Milton Model | |
|---|---|---|
| What it does | Makes language specific | Leaves selected details open |
| Information | Recovers what was left out | Lets the listener supply what is missing |
| Assumptions | Questions them | Introduces new ones, gently |
| Purpose | Clarify problems and outcomes | Encourage imagination and exploration |
| Form | Direct questions | Permissive suggestions |
| Sounds like | “What specifically stops you?” | “You may discover what could help you move forward.” |
| Failure mode | Interrogation | Waffle |
| Direction | Towards precision | Towards possibility |
One Statement, Two Treatments
“I can’t lead confidently.”
Meta Model Response
“What specifically stops you from leading confidently?”
Investigates the limitation. Recovers the obstacle by name.
Milton Model Response
“You may begin to discover that confidence develops differently in different situations.”
Creates space around the limitation. Loosens it before it is examined.
Neither response is superior. The skill is knowing which one the moment is asking for — and that judgement is what separates a practitioner from someone who has memorised a list. Read the other half: the NLP Meta Model.
The Framework
Grouped by what they do for you, rather than by academic category. Learn one at a time. One pattern used well beats five patterns crammed into a sentence — which is how most people give themselves away.
| Pattern | Example | Watch Out For |
|---|---|---|
| 1. Pacing Current Experience Describe what is verifiably happening. | “As you read these words, you may begin to notice which examples are most relevant to you.” | Pace what is true. Pretending to know their experience destroys trust instantly. |
| 2. Pacing And Leading Current experience + possible direction. | “You have been thinking carefully about this move, and you may now be ready to consider what matters most.” | Leading too early. A person who does not feel understood will not follow you anywhere. |
| 3. Truisms Broad statements most people accept. | “Every meaningful change begins somewhere, and your first step does not have to be perfect.” | Presenting your opinion as a universal truth. |
| 4. Utilisation Use whatever shows up, including resistance. | “That uncertainty may help you examine the idea carefully and decide what would make it useful.” | Nothing. This is the single most valuable pattern here. |
| 5. Simple Conjunctions And, while, as, when. | “You can review the options and begin noticing which one feels more practical.” | Implying a connection that does not exist. |
| 6. Cause And Effect One experience can lead to another. | “As you organise your thoughts, you may find it easier to communicate your decision.” | Say “this may help”, never “this will definitely make”. |
| 7. Complex Equivalence Give an experience a useful meaning. | “Feeling nervous may mean this opportunity matters to you.” | Offer the meaning. Never impose it as the only one. |
| Pattern | Example | Watch Out For |
|---|---|---|
| 8. Simple Deletion Leave the detail out. | “You can become more confident.” — about what? when? how? | Deleting so much that nothing lands at all. |
| 9. Unspecified Referential Index Who is never named. | “People can learn to communicate differently.” | Hiding behind “people” when you mean one specific person. |
| 10. Unspecified Verbs Learn, change, discover, realise, notice. | “You may discover a better approach.” | The missing “how” is the gift. Do not fill it in for them. |
| 11. Nominalisations Process frozen into a noun. | “Greater clarity can create movement.” | Overuse produces sentences that sound profound and say nothing. |
| 12. Unspecified Comparisons Easier, better, more comfortable — than what? | “You may find an easier way.” | Where evidence or accountability matters, state the comparison. |
| 13. Universal Quantifiers Every, all, always, everyone. | “Every conversation creates an impression.” | Use sparingly. Broad claims are easy to challenge, and one challenge costs you the room. |
| 14. Modal Operators Can, could, may, might. | “You may begin developing confidence in your own way.” | Possibility opens the door. Necessity closes it: avoid “you must be confident”. |
| Pattern | Example | Watch Out For |
|---|---|---|
| 15. Presuppositions Something must be assumed for the sentence to work. | “Which part will you practise first?” — assumes practice; only the order is open. | Never presuppose away a choice the person actually needs to make. |
| 16. Embedded Suggestions A direction placed inside a larger sentence. | “Some people find they can speak with real confidence once they focus on serving the audience.” | These are not secret commands. Effect depends on trust, timing, relevance and willingness. |
| 17. Conversational Postulates A question that functions as a request. | “Could you take a moment to reflect on that?” | Nothing. This is the most socially natural pattern in the set. |
| 18. Tag Questions A short question appended to a statement. | “Clear expectations make teamwork easier, don’t they?” | Frequent use sounds pushy or insecure. Invite reflection; never press for compliance. |
| 19. Double Binds Two choices, both leading the same way. | “Would you prefer to discuss the goal first, or begin with the challenge?” | An ethical double bind leaves a real third option, including saying no. Remove that and it is a trap. |
| 20. Negative Suggestions Mentioning an action while negating it. | “You do not have to decide immediately.” | Heavily overstated in NLP teaching. People understand negatives perfectly well. Where safety matters, be direct. |
| 21. Extended Quotes The suggestion sits inside someone else’s words. | “A leader once told me her mentor used to say: you do not have to know everything before you take the next thoughtful step.” | If the listener loses the thread, influence becomes irritation. |
| Pattern | Example | Watch Out For |
|---|---|---|
| 22. Metaphor Explain one situation through another. | “Old habits are like well-used roads. They feel natural because we have travelled them so often.” | A metaphor must fit their experience, not showcase yours. |
| 23. Storytelling Carry the idea inside a narrative. | “I worked with a leader who believed confidence had to come before action. She found it arrived after several small ones.” | Stories turn manipulative the moment they hide facts, invent proof or work through fear. |
| 24. Ambiguity More than one possible reading — phonological, syntactic, scope. | “You can comfortably review your options and make a decision.” — comfortably reviewing, or comfortably deciding? | Some ambiguity invites interpretation. Too much produces confusion. |
| 25. Selectional Restriction Violation Give a human quality to something that has none. | “That old habit may be ready to retire.” | It should sound natural, never theatrical. |
Interactive Tool
Five steps, one sentence. Give it the outcome and the person's current reality, and it assembles a Milton Model sentence you can actually say out loud — then shows you which patterns you used.
Step 1
What would you like this person to consider? Not to do — to consider. If you cannot name it, stop here; a sentence without an outcome behind it is word salad.
Step 2
Something they already know to be true. This is your pace — and it must be accurate. A false pace is worse than no pace at all.
Step 3
This is the modal operator of possibility. It is what stops the sentence being an order.
Step 4
One pattern. Not four. This decides the shape of the second half of the sentence.
Fill in the outcome and the pace first — those two carry the sentence.
Your Sentence
Patterns You Used
Read it back. Does it sound like something a human being would say to another human being? If it sounds like a hypnosis script, take a pattern out. The Milton Model should make communication more human, never more mysterious.
This is what will be sent:
Your sentence is private to you — nothing was stored on this page. Sharing it is entirely your choice.
Step By Step
What do you want the person to consider — and is it useful for them? Does it preserve their choice? A Milton Model sentence with no outcome behind it is decoration. Name the outcome before you open your mouth.
Start with what is verifiably true. “You have been thinking about this change for several weeks…” Accuracy is the whole value here. A pace the person does not recognise costs you the sentence and some of the relationship.
“…and you may begin to notice…” Possibility rather than instruction. Nothing to comply with, so nothing to resist.
“…which first step would make the decision easier.” Open, and pointed. The sentence does not say which step. It insists there is one.
Stop talking. The listener now decides what the first step is — and because they chose it, they will act on it. Filling the space yourself undoes everything the previous four steps built.
The finished sentence: “You have been thinking about this change for several weeks, and you may begin to notice which first step would make the decision easier.” One pace, one permissive phrase, one presupposition, one open ending. Four moving parts, and not one of them announces itself.
Interactive Tool
The same intent, delivered two ways. One produces resistance. One produces reflection. Move the switch and watch what changes — and what stays exactly the same.
Notice what did not change. The speaker wants the same thing in both versions. The Milton Model is not a softer opinion — it is the same opinion, delivered so the listener can arrive at it themselves. If your outcome would embarrass you to state plainly, no amount of permissive language makes it ethical.
The Ethics Gate
Every tool of influence can be used well or badly, and this one leaves fewer fingerprints than most. Four questions. Answer all four honestly before you open your mouth.
What useful outcome does this support — for them, not only for you? Can you say it out loud without flinching?
What is genuinely true about their present experience? Are you pacing something real, or something you wish were real?
Can this person still disagree, pause or say no? A double bind with no third door is not a choice — it is a corridor.
Can you say it naturally, in your own voice, without sounding like a badly written hypnosis script?
Four unticked. Work through them honestly. The patterns are easy; this part is the job.
The Milton Model becomes manipulation the moment it is used to hide important information, manufacture urgency or fear, pressure agreement, disguise a command as a choice, override a boundary, promise what you cannot evidence, or work on someone's vulnerability.
Responsible influence is transparent about its outcome. It respects the other person's interests, their intelligence and their freedom to choose. The patterns do not make you ethical. Intention, transparency and respect do.
Where It Earns Its Place
A Result From The Field
Mercedes-Benz trained its customer-facing teams in NLP language and influence — pacing the customer's reality before leading anywhere, and replacing instruction with invitation. Customer satisfaction moved from 72% to 96%, and the annual targets were met in eight months. Resistance, it turned out, had been a language problem.
Direct: “You need to stop overthinking and take action.”
Milton: “As you consider what has been keeping you stuck, you may begin to recognise one small action
that feels possible now.” The client arrives at the action. That is why they do it. Trained formally in
ICF coach certification.
Direct: “Your team must accept the new strategy.”
Milton: “As the team understands why the strategy is changing, people may begin to see where they can
contribute.” Ownership cannot be issued. It has to be found. Core to
leadership development.
Direct: “This is the best option, so you should buy it.”
Milton: “As you compare the options against your priorities, you can decide which one gives you the
strongest business value.” Good selling supports a decision. It never disguises one. See
NLP sales training.
Direct: “Remember everything I am teaching.”
Milton: “You may not consciously recall every point today, and the ideas that matter most can return
when you need to apply them.”
Feedback delivered as verdict produces defence. Delivered as an invitation to look, it produces looking. “As you think back over how that meeting went, which moment would you handle differently?”
“I’m not good at public speaking” becomes “I don’t yet communicate as confidently as I want, and that can change with preparation and practice.” The word yet is doing all the work, and it costs nothing.
The Other Half Of The Skill
This section matters more than the twenty-five patterns. Artful vagueness applied to the wrong subject is not influence — it is negligence, and occasionally it is a breach of contract.
Be Clear, Direct And Specific About
“You may discover the right time to complete the report.”
“Please submit the report by 4:00 PM on Friday.”
“Do not imagine touching the machine.”
“Keep your hands away from the machine.”
“The investment may feel comfortable in time.”
“The fee is X, payable on the first of the month.”
Precision and influence are not enemies. The right language depends entirely on the situation — and knowing which situation you are in is the skill that takes longest to build. For the other direction, see the NLP Meta Model.
Refining The Skill
Cause: The belief that slow speech and dramatic pauses do the work. Fix: Speak normally. Every pattern here works better in your own voice than in a borrowed one.
Cause: Enthusiasm. A sentence carrying a presupposition, a double bind, an ambiguity and an embedded suggestion sounds absurd. Fix: One well-chosen pattern beats five forced ones.
Cause: Nominalisations with nothing underneath. “Transformation creates empowerment through alignment” sounds impressive and says nothing. Fix: Attach every abstract word to a real situation the person recognises.
Cause: Guessing at their state. “You are becoming excited about this proposal” — they may well not be. Fix: Pace only what is observable. “As you consider the proposal, you can notice which parts are most relevant.”
Cause: Believing the phrase does the work. Fix: No phrase guarantees agreement. Trust, timing, context, tone, relationship, credibility and the quality of the idea decide the outcome. The words are the smallest part.
Cause: Reaching for artful vagueness with nowhere to send it. Fix: Know why you are speaking before you decide how. Vagueness without direction is word salad, however beautifully phrased.
Twenty Questions Answered
Every question, answered in full. Filter by what brought you here — or read all twenty in order.
The Milton Model is a set of language patterns that deliberately leaves some information open, so listeners connect the message to their own thoughts, experiences and meanings. Instead of “be confident”, it offers “you may begin to notice confidence growing in your own way.” The first tells someone what to feel. The second gives them room to find it.
It is named after Dr Milton H. Erickson, an American psychiatrist known for medical hypnosis and psychotherapy. Erickson worked through story, metaphor, indirect suggestion and permissive language rather than direct instruction. Richard Bandler and John Grinder studied his communication and set out the recurring patterns in Patterns of the Hypnotic Techniques of Milton H. Erickson, M.D. The model is an NLP interpretation of those patterns — not a complete account of everything Erickson did.
Because the patterns were drawn from Erickson's hypnotic and therapeutic work: indirect suggestion, metaphor, ambiguity and permissive phrasing. The label describes the origin, not a guarantee of effect. The same patterns appear routinely in ordinary coaching, teaching, storytelling and persuasive speech, where nobody is being hypnotised at all.
Vagueness with a direction. “Something good will happen” is merely vague — it has no purpose and gives the listener nothing to do. “As you think about your next step, you may recognise an opportunity you had not considered” is artfully vague: still open, and yet it aims attention at the future, at possibility, at discovery. The listener decides what the opportunity is. The sentence insists there is one.
“As you think about the situation, you may begin to recognise a possibility that was not obvious before.” It paces the act of thinking, uses permissive language, and leaves the exact possibility open. Three patterns, and none of them announces itself.
Start with pacing and leading, permissive language, presuppositions, conversational postulates, metaphors, unspecified verbs and embedded suggestions. These are the ones that sound natural the first time you use them. Ambiguity and extended quotes are considerably harder to use without sounding strange, and they can wait.
Pacing is describing what is verifiably true right now. Leading is introducing a possible next step. The formula is: current experience, then possible direction. “You have already identified what is not working, and the next conversation may help you discover what could work instead.” Lead too early, before the person feels understood, and they will follow you nowhere.
Two choices that both lead in the same direction. “Would you like to practise this today or tomorrow?” Both options assume practice. An ethical double bind leaves a genuine third option intact — including the right to say no. A double bind that removes the real decision while presenting the appearance of choice is not influence. It is a corridor with the doors painted on.
Not as secret commands, and anyone selling them that way is overstating the case. An embedded suggestion places a direction inside a larger sentence, sometimes marked by a slight shift in pace or tone. Whether it lands depends on context, trust, delivery, relevance and the listener's willingness — the same things that decide whether any sentence lands. There is no phrase that bypasses a person's judgement.
They are best described as complementary opposites. The Meta Model makes vague language specific by asking precise questions. The Milton Model deliberately leaves selected details open so the listener supplies their own meaning. One recovers information; the other invites it. A capable communicator moves between them several times in a single conversation.
Given the statement “I cannot lead confidently”, the Meta Model asks “what specifically stops you?” — it investigates the limitation. The Milton Model offers “you may begin to discover that confidence develops differently in different situations” — it creates space around the limitation. The Meta Model moves towards precision. The Milton Model moves towards possibility. Neither is superior; they answer different moments.
The Meta Model, without much doubt. It is more systematic, it builds the listening that everything else in NLP depends on, and its failure mode — sounding like an interrogator — is obvious enough that you will correct it quickly. The Milton Model's failure mode is producing pleasant-sounding sentences that accomplish nothing, and you can do that for years without noticing.
No. A single Milton Model sentence does not hypnotise or control anyone. Hypnosis is a broader process involving attention, expectation, context, responsiveness and consent. These same patterns appear constantly in ordinary coaching, teaching, storytelling and advertising, where nobody imagines a trance is taking place. Anyone who tells you these words grant control over other people is selling something.
Any communication tool can be used responsibly or irresponsibly, and this one leaves fewer fingerprints than most. It becomes manipulation when it is used to hide important information, manufacture urgency or fear, pressure agreement, disguise a command as a choice, override a boundary, promise what cannot be evidenced, or work on someone's vulnerability. Responsible influence is transparent about its outcome and respects the other person's interests, intelligence and freedom to choose.
Influence respects informed choice. Manipulation hides the real intention, or works through pressure, deception or emotional exploitation. The practical test is simple: could you state your outcome out loud to the person, and would they still be willing to continue? If yes, you are influencing. If the answer makes you uncomfortable, no amount of permissive language repairs it.
The Milton Model is taught as a communication framework derived from an interpretation of Erickson's work. It should not be presented as a scientifically established method for controlling behaviour or treating health conditions, and reviews of NLP's clinical claims have found the evidence limited. What it offers is a practical way of speaking that lowers resistance and invites reflection — useful in coaching, leadership, sales and teaching, and no substitute for clinical treatment where clinical treatment is what is needed.
Yes — to encourage reflection, introduce change, communicate a direction and invite ownership. Ownership cannot be issued as an instruction; it has to be arrived at. But for expectations, deadlines, accountability and performance standards, leaders should be clear, direct and specific. A leader who is artfully vague about a deadline is not influencing anyone. They are being unclear.
It can help a salesperson ask thoughtful questions, invite the customer to imagine an outcome and open up possibilities they had not considered. It must never be used to obscure conditions, manufacture urgency or pressure a decision. Good sales communication supports a decision; it never disguises one. The moment a buyer feels handled rather than helped, you have lost more than the sale.
Yes — permissive questions, metaphor and open language help clients explore possibilities they would defend if handed to them. The professional line is this: a coach should not plant their own answer and then pretend it came from the client. That is not coaching. It is ventriloquism with better manners.
The patterns can be understood in an afternoon. Using them naturally takes listening, practice, feedback and judgement. Memorising twenty-five names is easy; knowing which one the moment calls for is the actual skill, and it is built by speaking, being heard, and being told honestly how it landed. Study one pattern at a time, write your own examples, use them in ordinary conversation, and ask someone to tell you whether you sounded like yourself.
Where To Go Next
The real skill is not sounding hypnotic. It is knowing when to be specific and when to leave space — and hearing, in the moment, which one the person in front of you needs. That is built with practice and honest feedback. Train both language models with an NLP Master Trainer on a certified NLP programme, or bring them to your team through corporate training.