There are two rooms.
In the first one, you’re excellent. You say a thing, people move. Your read on the situation is right, your instincts fire correctly, and afterwards you don’t even think about how it went because it obviously went fine.
In the second room, you might as well be broadcasting on a frequency nobody’s tuned to.
Same you. Same preparation — often more preparation, because you know it’s the difficult room. Same quality of thinking.
And half of what you say lands somewhere on the table between you and never gets picked up.
You’ve explained this to yourself. Everybody does. It’s politics, or he’s threatened, or they’ve already decided, or there’s history you’re not party to.
Sometimes that’s true.
But run the numbers on it honestly: it’s the same room every time. The same person, or the same kind of person. If it were politics you’d expect it to move around. It doesn’t. It’s a fixed, repeating, entirely predictable failure with one specific category of listener, and that’s not a description of a conspiracy.
That’s a description of a mismatch.
What’s Actually Failing
Persuasion fails most often not because an argument is weak but because it is delivered in a register the listener cannot process. Every message carries two components: the content, which is the argument itself, and the register, which is the type of value the argument is framed in — evidence, outcome, relationship, or precision. A listener attends to the register they operate in and discounts the others, largely without noticing they are doing it. When register and listener match, the content is received and evaluated. When they mismatch, the content is not rejected on its merits; it is not fully received at all, and the listener experiences the exchange as vague, irrelevant, or excessive. Because most speakers default to the register they personally find persuasive, mismatch failures cluster around specific listener types and recur.
That’s the mechanism. Now the part that’s actually about you.
You Only Have One Register
Here’s the uncomfortable bit.
You think you adapt. Most people do — you’d say you talk differently to the board than to the team, differently to a client than to a supplier.
What you’re adapting is tone, formality, and length. Almost nobody adapts register, because register doesn’t feel like a choice. It feels like what makes something convincing.
And what makes something convincing to you is a fact about you, not about persuasion.
There are four common ones. Read them and you’ll know yours inside about ten seconds, which is itself worth noticing, because a thing you can identify that fast is a thing you’ve been doing on rails.
Evidence. You lead with the case. Data, precedent, worked examples, the objections pre-handled. You believe a well-supported argument is intrinsically more persuasive than a poorly-supported one, and you find it slightly obscene that this isn’t universally true.
Outcome. You lead with consequence. What this moves, what it prevents, what the world looks like afterwards. You skip the method because the method is boring and the method isn’t the point.
Relationship. You lead with rapport. You build the connection first and the content second, because in your experience people say yes to people rather than to proposals, and you’re not wrong about that with a large fraction of the population.
Precision. You lead with definition. You want the terms agreed, the scope bounded, the ambiguity removed. You genuinely can’t proceed while a key word is doing two jobs, and you assume nobody else can either.
Pick yours. Then notice: it’s the same one you use everywhere.
That’s the whole problem, in one line. You are running a single carrier signal across an entire population of receivers, and you’ve been evaluating the results as though the variable were the message.
The Carrier and the Payload
The model I find most useful here is transmission.
Your argument is the payload. The register is the carrier.
You’ve spent enormous effort on the payload — the logic, the evidence, the anticipated objections, the neat closing line. All of that is real work and none of it is wasted.
But the payload only unloads if the carrier is one the listener receives.
Wrong carrier, and something specific happens that’s worth understanding precisely: the listener doesn’t disagree with you. Disagreement would be a much better outcome. Disagreement means the content arrived and was assessed.
What actually happens is vaguer and worse. They nod. They say something non-committal. There’s no argument, no pushback, nothing to grip. And later, when the decision goes elsewhere, nobody can tell you what was wrong with your case — because nothing was wrong with your case. It didn’t get evaluated.
That’s the signature of a register failure, and once you know it you’ll spot it everywhere: the absence of engagement rather than the presence of objection.
If people argue with you, your register is landing. If people nod and nothing happens, it isn’t.
Why the Worse Version Won
Now the thing that genuinely stings, and I’d rather name it than let you keep explaining it away.
Two weeks after your proposal died, somebody presented a worse version and got it through.
Less prepared. Less rigorous. Possibly wrong in two places you could identify from across the room.
Here’s what happened, and it isn’t politics: their default register happened to match the listener’s.
That’s it. They weren’t better. They were lucky, in the specific and slightly infuriating sense that the one register they own was the one that room takes.
Which means their success is as unreliable as your failure. Put them in your good room and they’ll die in it, and they’ll go home and tell themselves it was politics.
Everybody in this story is running one carrier and attributing the results to merit.
The Read, and the Adjustment
So what do you actually do about it. Two moves.
First: read what they lead with.
You don’t need a personality assessment. You need about two sentences, because people reveal their register in what they ask about first.
The listener who asks how do we know that? is on evidence.
The listener who asks what does this get us? is on outcome.
The listener who asks whose idea was this, who else is involved? is on relationship.
The listener who asks what exactly do you mean by that? is on precision.
That’s it. First substantive question, and you have your read.
If they haven’t asked anything yet, look at what they treat as too obvious to mention — people skip the layer they’re fluent in.
Second: lead with theirs, not yours.
Not translate everything. Not abandon your argument. Just change the first ninety seconds, because the first ninety seconds decide whether the rest gets processed or filed.
The evidence-first speaker talking to an outcome listener leads with the consequence and holds the data in reserve for when it’s asked for.
The outcome-first speaker talking to a precision listener defines the terms before making any claim, however impatient that feels.
And here’s the counterintuitive one, which catches thorough people every time: with certain listeners, less evidence performs better than more. Bring a large stack of support to someone who argues for sport and you’ve handed him more surface to attack — every additional claim is another handhold. That failure mode has its own piece: why the worse communicator wins.
What This Is Not
It is not “know your audience.” That advice is true and useless, because it tells you to do something without telling you what varies. The variable is register, it’s one of about four, and it’s readable from a single question. That’s actionable in a way “know your audience” never has been.
It is not manipulation. You’re not changing what you’re arguing for. You’re changing the carrier so the argument arrives intact. The test is simple: would the adjustment still work if the listener knew you were making it? Framing a proposal in outcome terms for a busy executive survives disclosure comfortably. Techniques that don’t survive disclosure are a different category, and they’re covered in reading people versus manipulating them.
It is not about likeability. Some of the worst register mismatches happen between people who like each other enormously. Warmth doesn’t carry content. Two people can have an excellent relationship and a persistent, decade-long failure to transmit anything of substance to each other, and usually neither of them can name why.
The Cost
Every piece here names the price. This one’s is more annoying than it sounds.
You have to stop leading with the thing you’re best at.
If your register is evidence, you’ve spent fifteen years building the skill of constructing an unimpeachable case. That skill is real, it’s rare, and it’s the thing you’re proudest of professionally.
And this tells you that with about half the people you’ll meet, the case should go second.
That feels like being asked to hide your best work. It feels like dumbing down. It isn’t — the case still gets made, and made properly, once the listener is in a state to receive it. But the sensation of leading with something you consider less rigorous is genuinely unpleasant and it doesn’t fully go away.
Second cost, smaller: you can no longer blame the room. Once you can see register, “he wasn’t listening” resolves into “I led with mine again,” and that’s a considerably less comfortable sentence to say in the car on the way home.
Common Questions
Why do my words work on some people and not others?
Because you’re delivering in one register and people receive in four. Matched register means your argument gets evaluated; mismatched register means it isn’t fully received, and the listener experiences the exchange as vague rather than as wrong.
How do I know what register someone uses?
Listen to their first substantive question. “How do we know?” is evidence. “What does this get us?” is outcome. “Who else is involved?” is relationship. “What do you mean by that?” is precision. One question is usually enough for a working read.
Why do my ideas get ignored when nobody argues with them?
Because the absence of objection is the signature of a register failure. Disagreement means the content arrived and got assessed. Silence and non-committal nodding usually means it didn’t arrive at all.
Is this the same as changing my personality to suit people?
No. Register is delivery, not position. You’re not adjusting what you argue for or what you believe, only which door you bring it through. If you find yourself changing the conclusion to suit the listener, that’s a different thing and it’s worth stopping.
What if I misread someone mid-conversation?
Change register and continue. Recovery is much cheaper than people expect — most listeners register an adjustment as attentiveness rather than as an error. What’s expensive is committing harder to a read that’s already failing.
Does this work upward, with people more senior than me?
It works best upward, because senior listeners have the least tolerance for a mismatched carrier and the least time to work through one. They’re also the most consistent, since they’ve been in the same register for decades.
Where to Start
One room. This week.
Pick the room you fail in — you already know which one — and before you go in, answer two questions on paper.
What register does this person operate in? One word: evidence, outcome, relationship, precision.
What am I currently leading with? One word, same list.
If they’re the same word, register isn’t your problem and you should look elsewhere. That happens and it’s useful information.
Nine times out of ten they’ll be different, and nine times out of ten yours will be the same word you’d have written for every other room in your life.
Then change the first ninety seconds. Only the first ninety seconds. Leave the rest of your material exactly as it is.
That’s the entire experiment, and it’s usually enough to convince you the full system is worth learning — which is Guiguzi’s nine types, where these four registers open out into nine listener states and the specific thing each one needs to hear.
I’ve spent twenty-five years working on how ordinary people develop capabilities that look impossible from the outside, and the person who can walk into any room and be heard is almost never the best-prepared person present.
They’re just the one who isn’t broadcasting on a single frequency and calling it clarity.