New NLP Practitioners ask me this question, more than almost anything else: “What’s the right way to structure an NLP session?”
And I get why. When you’re new, the temptation is to reach for a script because it feels safe. But in real sessions, scripts make you clumsy. They pull your attention off the client and into your head. They make you miss the very thing you’re there to track: how this specific person is running their experience.
So when I talk about “structure,” I’m not talking about a memorized sequence of techniques. I’m talking about a container — a simple rhythm that keeps you oriented while you stay fully present.
Start by setting the frame
With a brand-new client, I almost always begin with a short pre-talk. Not because I love pre-talks, but because they prevent unnecessary mess.
A good pre-talk does something subtle: it tells the client’s nervous system, “You’re safe here. This will be useful. You don’t have to perform.” It also quietly protects the work from the client’s habitual meaning-making. Once in a while you’ll meet someone whose mind automatically scans for failure — the internal question is essentially “What if this doesn’t work?” And when that question is running, it’s like coaching with a brake half-pressed.
So I’ll normalize a few things up front. That change can be quick, but it can also unfold over days. If something doesn’t feel “done” in the room, that doesn’t mean it failed — it may mean their system is still reorganizing. And we’ll measure success by what happens in real life, not by whether they can describe the change perfectly in the moment.
With an ongoing client, the pre-talk turns into a check-in. I want to hear what’s happened since last time — what improved, what held, what surprised them, and where the old pattern tried to reassert itself. It’s rare that a continuing client reports “that didn’t work.” What I hear far more often is: “It worked… until a new situation showed up.” That’s not failure. That’s simply the next layer of generalization. That’s the work.
I’ll also zoom out briefly and revisit longer-range outcomes. Over time, goals drift. Sometimes in a good way. Sometimes in a way that quietly derails momentum. A quick check for scope creep keeps the coaching arc clean.
Then we get ruthless about outcomes
After that initial contact, we clarify the outcome for the session — and we do it in a way that makes it achievable.
People don’t realize how much their language shapes their results. If someone states an outcome as a vague emotional wish or a complaint disguised as a goal, they’re making their own change harder. So I tighten it. I want to know what they want instead, how they’ll know they’re getting it, and where they want it to show up first.
This is also where I calibrate the “bar” for the session. Very occasionally someone sets expectations unrealistically high for a single meeting — it happens, but it’s rare. Much more often, they set the bar painfully low: “Help me get through this one unpleasant conversation.”
And I’ll say, gently but directly: Why would we settle for one conversation? Why not upgrade the pattern so it changes this conversation and every similar one from now on? That’s the difference between patching a moment and building capability.
Another thing I watch for here is the client’s attachment to a particular method. Sometimes a client will come in requesting a specific technique — and sometimes that’s fine. But sometimes it’s a clue that they’ve chosen a means from a very limited view of what’s possible. So I’ll interview my way to the end result first. I care deeply about the ends. I’m not necessarily invested in the client’s chosen means.
Now the part that matters: stop using scripts
This is where I’m going to be blunt.
If you’re using a script in a live NLP session, you’re probably missing most of what you need to notice.
Human communication is dense. Even when people are trying to be unclear, they give you an enormous amount of information right away — verbally and nonverbally, in parallel. Their first few sentences often reveal how they’re structuring the problem, what they’re attending to, what they’re deleting, and what kind of solution will fit them.
If a practitioner isn’t seeing, hearing, and feeling enough information early in the conversation, the answer is not “find a better script.” The answer is: train sensory acuity. Train calibration. Train pattern literacy.
Because the real skill is not “running Technique X.” The real skill is being present enough to recognize what’s happening and to design something that matches this person’s system.
That design might involve a classic NLP intervention or technique. It might involve habit-building, a pattern interrupt, timeline work, reframing, hypnotic language, strategy installation and the art of strategy pipelines, state shifts, whatever fits.. But the more “cookie-cutter” it gets, the less effective it tends to be. People can feel when they’re being processed through a template. And their unconscious mind often resists being treated like a form you fill in.
The session becomes a series of tests, not a performance
Once the outcome is clear and you’ve got a sense of the pattern, the work tends to move like this: you make a move, you test, you adjust.
Does their state change? Does their perception shift? Do they generate better options without you spoon-feeding them? Can we bring up the old trigger and have it land differently? Can we future pace it into the hard context?
If something doesn’t land, I don’t grind. Grinding is usually a sign you’re trying to force a solution that doesn’t match. I’ll shift levels — environment, behavior, capabilities, beliefs, identity — or change angle entirely. One of the most underrated coaching skills is knowing when to pivot quickly.
Then I task — and I mean real tasking
At some point, I want the change to have legs outside the room. That’s where tasking comes in.
Most people interpret “homework” as a basic follow-up assignment. Ericksonian tasking can be that, but it can also be far more elegant. It’s the art of designing something that meets the client’s conscious mind where it’s comfortable, while quietly recruiting the unconscious mind to organize new learning in the background.
Sometimes that’s straightforward: practice a behavior, run a pattern, gather evidence. Sometimes it’s metaphorical. Sometimes it’s a task they won’t consciously understand — yet over time it creates exactly the change they came in for.
Tasking does more than reinforce progress. It installs identity through action. And it also communicates something important: you’re thinking beyond the hour. You’re thinking in terms of process — not just a clever intervention.
Wrap up based on the kind of work you did
Finally, I close the session in a way that supports consolidation.
If the work has been mostly conscious and explicit — lots of discussion, lots of insight — I’ll invite them to articulate what shifted. Not because I need praise, but because naming the change often locks in certainty. And I’ll phrase questions in a way that naturally evokes agreement and confidence.
But if the work has been more unconscious — hypnotic language, indirect suggestion, deeper internal reorganization — I’m careful about not provoking an early “no.” Not because the work failed, but because their system may need time, and the conscious mind loves to evaluate too soon.
In those cases I’ll ask open-ended, future-oriented questions that direct attention toward success without demanding a verdict: “I’m curious where you’ll notice this showing up first.” “In how many situations might you surprise yourself with this being easier?” That kind of language lets their unconscious mind do what it needs to do without the conscious mind trying to score the session on the spot.
Then I give a simple follow-through: what to do next, what to notice, what to do if old triggers show up, and any tasking I assigned. That last part matters. It helps them handle the unknown with comfort while the new pattern stabilizes.
The “structure” in one sentence
If you want the whole thing distilled, it’s this:
You frame the work, you clarify outcomes, you track patterns with real calibration, you design interventions responsively, you task for real-world installation, and you close in a way that protects consolidation.
Everything else is just style.


0 Comments