An NLP coaching session structure is a fixed sequence of six phases — contact, outcome, information, intervention, ecology, test — that the coach runs in the same order every time, while the content changes with each client. The order is the point: it stops you intervening before you know what “better” would look like, and it stops you ending a session without checking whether anything actually changed.
Below is the template we teach, with timings, the question that opens each phase, and the signal that tells you the phase is finished.
| Phase | Time (90 min) | Question that opens it | Done when |
|---|---|---|---|
| 1. Contact and baseline | 5–10 min | “How has the week been?” | You can describe how this person looks and sounds at rest |
| 2. Outcome | 15–20 min | “What do you want instead?” | The client states something you could both notice happening |
| 3. Information | 20–25 min | “How do you know?” | You can describe the pattern in sequence, not in adjectives |
| 4. Intervention | 15–20 min | “Let’s try something.” | The client’s state visibly changes while thinking of the same thing |
| 5. Ecology | 5–10 min | “Who else is affected by this?” | No part of the client’s life objects to the change |
| 6. Test and future pace | 10 min | “Imagine next Tuesday.” | The old response does not return under rehearsal |
Why does an NLP coaching session need a fixed structure at all?
Because each phase produces the raw material the next one needs, and skipping one leaves the following phase guessing. You cannot calibrate a change without a baseline. You cannot choose a technique without a defined outcome. You cannot test a result you never specified.
A fixed structure also does something less obvious: it protects the client from the coach’s enthusiasm. Newly certified practitioners have a handful of techniques they are excited to use, and without a structure the session bends towards the tool rather than the problem. The sequence forces twenty minutes of listening before anything gets “done” to anyone.
The structure is not a script. Within each phase you improvise freely — the phases simply tell you what job you are currently doing, which is surprisingly easy to lose track of in the middle of an emotional conversation.
What does each of the six phases actually do?
Each phase has one job and one exit condition. Here is what you are doing in each, and the mistake that most often happens there.
- Contact and baseline. Small talk with a purpose: you are registering posture, pace, breathing, volume and the words this person uses when nothing is being worked on. Everything you notice later is measured against this. Common mistake: starting the real work while still taking your notes out.
- Outcome. You turn a complaint into a well-formed outcome: stated positively, testable through the senses, within the client’s control, and located somewhere specific in their life. Common mistake: accepting the first answer. “Less anxious” is a direction, not an outcome.
- Information. Precision questions recover what ordinary language leaves out — who exactly, compared to what, what stops you, how do you know. You are after the sequence: what happens first, what follows, where the state flips. Common mistake: collecting history instead of structure.
- Intervention. One technique, applied properly. Anchoring, a submodality change, a reframe, parts work, perceptual positions. Common mistake: three techniques run loosely because the first one did not produce fireworks in ninety seconds.
- Ecology. Does the change fit the rest of this person’s life — their partner, their team, whatever the old pattern was quietly protecting? Common mistake: skipping it because the client is smiling.
- Test and future pacing. The client mentally rehearses the situation that used to trigger the pattern while you watch for the signals you calibrated in phase one. Common mistake: treating a good feeling in the room as evidence.
How do you fit this into 60 minutes instead of 90?
You cut the intervention, not the framing. In a 60-minute session the realistic split is roughly 5 / 15 / 20 / 10 / 5 / 5, which means the technique gets ten minutes and has to be chosen well rather than explored.
That sounds backwards to most new coaches, who assume the technique is the session and the questions are preamble. In practice a well-framed problem often changes during phase three, before any technique is applied, because the client hears their own pattern described in sequence for the first time. A badly framed problem cannot be rescued by a longer intervention.
If the outcome is still unclear at the 25-minute mark of a 60-minute session, the honest move is to spend the whole hour on framing and agree the intervention for next time. Clients rarely mind. They mind being run through a technique aimed at the wrong target.
How is a first session different from a follow-up?
A first session spends most of its time in phases one to three, and often ends without a formal intervention. A follow-up opens with a test of the last change and can reach the intervention within twenty minutes, because the baseline and the outcome already exist.
| First session | Follow-up session | |
|---|---|---|
| Opens with | Contracting: scope, confidentiality, what coaching is and is not | Testing what happened since |
| Bulk of the time | Outcome and information gathering | Intervention and ecology |
| Typical ending | An agreed outcome and one small experiment | A tested change and a rehearsal |
| Biggest risk | Intervening on the presented problem, not the real one | Assuming last week’s outcome is still the current one |
Contracting in the first session is not administrative box-ticking. It is where you say plainly that coaching complements professional care and does not replace it — that trauma, diagnosed anxiety or depressive disorders, addiction, eating disorders and anything involving risk of self-harm belong with a licensed psychologist, psychiatrist or doctor, and that you will refer if that is what shows up. The International Coaching Federation treats recognising those limits as a core competency rather than a failure, and saying it out loud in minute five is far easier than discovering it in minute fifty.
What should you write down between sessions?
Four things, and no transcript. Notes exist so that the next session can start at phase six rather than phase one, which means you record what you would otherwise have to re-elicit.
- The outcome, in the client’s own words. Paraphrasing it into your language is how outcomes quietly drift into what you think they should want.
- The baseline. Two or three concrete observations from phase one — how they sit, how fast they talk, the phrase they repeat. Useless-looking notes that become the measuring instrument.
- The structure of the pattern. The sequence you recovered: trigger, internal representation, state, behaviour. This is the part clients cannot reproduce for you later.
- What you did and what changed. Technique, and the observable difference — not “good session”, which tells your future self nothing.
Keep them somewhere that respects data protection law in your jurisdiction, and keep them shorter than you think. Notes that take twenty minutes to write get skipped by week three.
Where does the structure usually break?
At the boundary between phases two and three, almost every time. The coach has an outcome that sounds good, feels the session should be moving, and starts gathering information about the problem instead of about its structure. Forty minutes of interesting conversation later, there is nothing to intervene on.
The questioning model that prevents this is the Meta Model, set out by Richard Bandler and John Grinder in The Structure of Magic in 1975 — the oldest part of the field and still the most practically useful. It is not a technique you run; it is how you listen during phase three.
In training rooms, the sessions that stall are almost never stalled by the wrong technique. They stall because the coach was explaining when they should have been asking. Newly certified coaches explain; experienced ones ask a short question and wait through the silence. The structure is largely a device for buying yourself that silence. — Carolin Mallmann, Licensed Trainer of The Society of NLP, Lisbon
If you are on the client side of this rather than the coach side, the structure is still worth knowing: it tells you what a competent session should feel like. That is how our 1:1 NLP coaching in Lisbon and online runs, starting with a free first conversation and no technique in sight.
Frequently asked questions
How long should an NLP coaching session be?
Sixty to ninety minutes is the standard range, with ninety giving enough room to frame the outcome properly and still intervene. Sessions shorter than an hour tend to force the coach into a technique before the outcome is clear. Longer than two hours, attention drops and the test at the end stops being reliable.
Can you change the order of the phases?
You can loop backwards but not skip forwards. Returning from information gathering to redefine the outcome is normal and often necessary. Jumping to an intervention before the outcome exists is the single most common structural error, because there is then nothing to test the result against.
What if the client’s outcome changes mid-session?
Follow the new one, and say so explicitly. Clients frequently arrive with a socially acceptable version of the problem and reach the real one during precision questioning. Continuing with the original outcome because it is written in your notes wastes the most valuable moment of the session.
How do you know the intervention worked?
The client thinks about the same trigger and the old response does not appear — checked against the baseline you took at the start, not against how pleasant the room feels. Verbal confirmation alone is weak evidence, because clients are cooperative. Rehearsing a specific future situation is the stronger test.
Does this structure work for group or team sessions?
The sequence holds, but the outcome phase takes far longer because a group has to agree what “better” means before anything can be tested. Interventions are usually lighter and more conversational, since you cannot calibrate eight people at once. Individual patterns that surface in a group setting are better picked up one-to-one afterwards.
Keep exploring
- How coaches use NLP with clients — the same hour seen from inside, including which techniques come up most often in real client work.
- NLP for entrepreneurs — our newest piece, useful if the clients you coach are founders rather than employees.
Carolin Mallmann — Licensed Trainer of The Society of NLP, trained by Richard Bandler. She has trained 300+ people and runs English-language NLP trainings in Lisbon in groups of no more than ten.


