An embedded command is a direct instruction buried inside a longer, ordinary-sounding sentence, marked by a subtle shift in tone on the words that matter. You say something completely normal on the surface, "I don't know when you'll relax and notice how good that feels," but underneath it, a command is riding along, unseen. It works because the sentence gets evaluated as a whole, while the command inside it slips past evaluation entirely.
I use this constantly, in coaching sessions, in casual conversation, in the scripts I write for hypnosis work. Once you can hear it in your own speech, you start noticing how much of ordinary conversation already has commands buried in it, most people just aren't doing it on purpose.
Why a bare command gets rejected
Say "relax" to someone who isn't already expecting an instruction from you, and their first move is often resistance. Not because they're stubborn, but because a direct command triggers the part of the mind whose entire job is to check incoming information against what it already believes. That part asks a fast, almost automatic question: who is this person to be telling me what to do? If there's no good answer, the command gets rejected, sometimes even provokes the opposite response. This is the same evaluator I've written about elsewhere, the mental filter that checks every incoming suggestion and blocks anything that doesn't match existing belief. A bare command hands that filter something obvious to catch. It's a single, isolated unit of language with nothing to hide inside. There's nothing else in the sentence for the filter to be busy with, so all its attention goes straight to judging the one thing you said.
Bury that same instruction inside a longer sentence, and the filter has a completely different job now. It's tracking the whole sentence, the subject, the verb, the meaning of the thing you're actually talking about. The command isn't sitting alone anymore waiting to be judged. It's riding inside content that reads as harmless, and by the time the sentence is finished, it's already landed.
How to construct an embedded command
Every embedded command has three parts working together. Miss one and it either doesn't land, or it's obvious enough that the filter catches it anyway.
The carrier sentence
This is the ordinary, harmless-sounding sentence that gives the command somewhere to hide. It needs to make complete sense on its own, with a topic that has nothing overtly to do with the instruction you're delivering. Good carrier sentences talk about time, uncertainty, or observation, "I wonder when," "you might notice," "some people find that," because those openers give you a completely natural reason to keep talking without raising suspicion.
The embedded phrase
This is the actual instruction, and it should be short. Two, three, maybe four words. "Relax now." "Trust me." "Let go." "Feel safe." The shorter it is, the easier it slides through the carrier sentence without disrupting its grammar or its meaning. If your embedded phrase is long enough to feel like its own sentence, you've built two commands stitched together, not one command hidden inside a sentence.
The tonal marker
This is what actually does the work of embedding it. Without a tonal marker, the words are just sitting in the sentence like any other words, no different from "the" or "and." The marker is what tells the subconscious, this part matters, pay attention here, while telling the conscious mind nothing at all because it's still parsing the sentence as one continuous thought.
The tone-shift technique
The marker itself is small. A slight drop in pitch, a small slowdown in pace, a touch more weight on the words as you say them, sometimes paired with the faintest pause right before the embedded phrase begins. You're not stopping the sentence and you're not changing your face or your posture. You're changing the sound of your voice for two or three words, then returning to your normal rhythm for the rest of the sentence. I think of it as underlining with my voice instead of my hand. If I read a sentence aloud and imagine which words I'd underline on the page, that's exactly where the tone shifts. The command gets a slightly different acoustic signature than everything around it, distinct enough for the part of the mind that's always listening underneath the words, invisible enough that the part of the mind judging the sentence for meaning never flags it as separate.
This only works if the shift is subtle. Exaggerate it and you've turned your embedded command back into a bare command with extra steps, now it's obvious, now the filter catches it, now you're back to square one. The goal is a marker just strong enough for the underneath-part of the mind to register, and just light enough that the surface conversation never breaks stride.
Why the conscious mind never isolates the command
This comes back to something I keep coming back to in everything I teach: the conscious mind processes in whole chunks, not word by word. When you hear a sentence, you're not evaluating each word in isolation and then assembling meaning afterward, you're taking in the whole shape of the sentence and extracting its overall point. That's simply how listening works, and it's the exact gap embedded commands are built to use. Because the evaluating mind is busy with the sentence as a whole, "I don't know when you'll relax and notice how good that feels" reads as one unit, a slightly poetic, slightly vague observation about you. It never stops to pull "relax" out of that sentence and ask whether it should obey it, because it was never presented with "relax" on its own. It was presented with the whole sentence, and the whole sentence is the thing it judged. Meanwhile, the part of the mind that isn't doing that evaluating, the part that's always listening underneath the conversation for what's relevant to it, picks up the tonally marked words and treats them as exactly what they are: an instruction. It doesn't need the sentence to make grammatical sense as a command, because it isn't parsing grammar. It's responding to the marker.
Example sentences
Here are four carrier sentences with the embedded command marked in bold. In each one, imagine your voice dropping slightly in pitch and slowing down just on the bold words, then returning to normal pace immediately after.
- "I don't know exactly when you'll relax completely, but I imagine it's sooner than you'd expect."
- "A lot of people notice that once they trust me, everything else gets easier."
- "You can think about it however long you need to, and at some point you'll just let go without even trying."
- "Some part of you already knows how to feel safe here, even if you haven't noticed it yet."
Notice that none of these carrier sentences would raise an eyebrow read on their own. Each one is the kind of thing a warm, thoughtful person might say in an ordinary conversation. That's exactly the point. The sentence has to be able to stand completely on its own, doing its own harmless job, while the marked words do a second job at the same time.
A closing word on using this well
Once you can hear the tonal marker in your own voice, this becomes something you can do in almost any sentence, and that's exactly why I want to be honest with you about it. This is a real skill with real effect, and like anything that bypasses a person's evaluating mind, it deserves to be used with care. Use it to help someone relax, to help someone feel safe, to guide someone toward something they already want for themselves, and you're doing exactly what this method is for. Use it to talk someone into something that isn't in their interest, and you've turned a tool for good into something else entirely. I teach this because I believe in what it can do when it's aimed at helping people. Aim it there, and only there.
Ready to actually use this?
The Black Book of Conversational Hypnosis gives you the exact words, in order, for real conversations, no talent required.
See the full system