How to be a better communicator: judge yourself by what the listener can repeat

Most advice on becoming a better communicator describes the communicator: be clear, be concise, be confident, listen actively, mind your body language. All true, and almost impossible to act on, because every one of those words is judged from inside your own head. You already think you were clear. Everyone does. The people who most need to improve are the ones least able to see it, because the feeling of having explained something well is produced by knowing what you meant, and you always know what you meant.

So this guide uses a different judge. A message was communicated well if the listener can say back, in their own words, the thing you needed them to take away. Not nod. Not say "makes sense." Say it back. That standard is harsh, measurable, and it turns out to be exactly the standard that the research on miscommunication keeps pointing at. It also happens to be the standard an AI speech coach or a recording can hold you to on your own, which matters, because nobody at work will volunteer that they did not follow you.

The guide covers spoken communication specifically: meetings, one-to-ones, updates, requests, feedback, disagreements. Writing gets a mention only where the research says the channel itself is the problem. It sits beneath how to improve speaking skills, which covers the delivery mechanics; this one is about whether the message arrives.

Here is the sequence:

1. Before you speak, write the one sentence you need the listener to be able to repeat.
2. Say that sentence first, then add only what changes their decision.
3. Replace every hint with the ask it stands for.
4. After an important exchange, check the playback: ask them to say what they are taking away.
5. When you are the listener, ask the question that moves their thinking rather than waiting silently.
6. Put anything with tone in it on a voice channel, not a text one.
7. Record one real conversation a week and score it on the five habits.

What does being a better communicator actually mean?

A better communicator is someone whose listeners end up holding the message that the speaker intended, more often and with less effort than before. Everything else, clarity, concision, warmth, confidence, is a means to that. It is useful to be blunt about this because the usual definition, a list of admirable qualities, lets you improve on every quality and still leave meetings where nobody knows what you wanted.

Think of a spoken message as passing through four states. There is what you meant. There is what you actually said, which is always a compressed version, because you leave out everything that is obvious to you. There is what the listener heard, which is filtered through what they already believe and whatever they were thinking about when you started. And there is what they can reproduce an hour later, which is a fraction of what they heard. Communication fails at every one of those joints, and speakers are poor at seeing any of them, because from the inside the four states feel like one.

The habit that separates good communicators from the rest is that they work on the last state and reason backwards. They decide what the listener must be able to repeat, and then they shape the saying to produce it. That is a different act from "explaining clearly," and it is why some people with ordinary vocabularies and plain delivery are easy to work with, while some polished speakers leave a room confused.

Why do you think you communicate better than you do?

Because you know what you meant, and that knowledge contaminates your estimate of what arrived. Psychologists call the mechanism egocentrism, and it is not a character flaw; it is the default setting of a mind that has only one perspective available at no cost, its own. Three studies make the size of the problem concrete.

In 2011, Kenneth Savitsky, Boaz Keysar, Nicholas Epley and colleagues sat 24 married couples, together an average of 14 years, in a circle facing away from each other and had each person speak ambiguous everyday phrases, "What have you been up to?", to convey one of four intended meanings. Speakers were confident their spouse would understand them better than the strangers in the circle. Their spouses did not: accuracy for spouses and strangers was statistically identical, and neither the speakers' nor the listeners' confidence predicted accuracy at all (University of Chicago News, 2011; Journal of Experimental Social Psychology, 2011). The authors named it the closeness-communication bias: with strangers we monitor what they might not know; with people we know well we let our guard down and talk as if they can see inside our head. Your long-standing colleagues are, for this purpose, your spouse.

In 2005, Justin Kruger, Nicholas Epley, Jason Parker and Zhi-Wen Ng ran a series of experiments on tone. In the first, senders expected 97% of their sarcastic or serious statements to be correctly decoded by email; 84% were. In the second, the same statements were either read by email or heard in the sender's voice. Listeners decoded nearly three-quarters of them. Readers were at chance, 50%. The senders were equally confident in both conditions (NYU Stern, 2005). The tone you hear in your own head while typing does not travel, and nothing in the act of typing tells you that.

In 2003, Roy Kessels reviewed the research on how much patients retain from a consultation. Between 40 and 80 percent of the medical information a clinician gives is forgotten immediately, and almost half of what the patient does retain is remembered incorrectly (Journal of the Royal Society of Medicine, 2003). These are patients with every reason to listen, receiving information about their own bodies from an expert they chose to see. A project update delivered to a colleague who is half-looking at a laptop does not do better.

50%

Source: Kruger, Epley, Parker & Ng, 2005

Put the three together and the picture is consistent. People you know well are not understanding you as well as you think. Tone is not surviving the channels you most often put it on. And whatever did arrive is decaying fast, with errors creeping into what is left. None of this is visible from the speaker's chair, which is why "be clearer" never works as an instruction. You need an external reading.

How to be a better communicator: the five habits

The five habits below each close one of the gaps in the diagram. They are ordered by how much they change what the listener can repeat, so if you take only one, take the first.

Habit 1: decide the playback sentence before you open your mouth

Before any exchange that matters, write or say to yourself the single sentence you need the other person to be able to repeat afterwards. Not the topic. The sentence. "We will miss the March date unless design signs off by Friday." "I need you to run the client call on Thursday without me." "The number in the deck is wrong and I am fixing it tonight."

This does two things. It forces you to find out whether you actually have a message, which a surprising share of meeting contributions do not. And it gives you the first sentence of what you are about to say, because the playback sentence should come first, before the context, before the reasons. Most people reverse this: they narrate the background so the conclusion will "make sense," and by the time the conclusion arrives the listener has formed their own, usually wrong, guess about where this is going. How to explain complex ideas simply covers the same move for technical material: say what the thing is for in one plain sentence, then add only what changes the decision.

A test for whether your sentence is a real playback sentence: could the listener act on it if they heard nothing else? If the answer is no, it is a topic, not a message.

Habit 2: say the ask, not the hint

The closeness-communication bias has one practical consequence that is worth a habit on its own. The better you know someone, the more you hint, because you assume they will fill in the rest. "It would be good to get the deck looked at before Thursday" is a hint. The ask it stands for is "Can you review the deck by Wednesday end of day so I can fix it Thursday morning?"

Hints feel polite and they feel efficient. They are neither, because the listener has to decode who is being asked to do what by when, and the research says their decoding is no better than a stranger's would be. The polite version of an ask is still an ask: a named person, a named action, a named time. If naming those feels too direct, soften the frame, not the content: "I know this is tight, and I would rather ask than assume: can you review the deck by Wednesday?"

A fast audit: listen back to a recording of yourself in a meeting and count the sentences that contain a request without a subject or a deadline. Those are hints. Each one is a message you believe you sent and the listener may not have received.

Habit 3: check the playback, not the nod

This is the habit that turns the rest of the guide from advice into measurement. Clinicians use a technique called teach-back: instead of asking a patient "do you understand?", which produces a yes from almost everyone, they ask the patient to explain the instructions in their own words, and then fix whatever came back wrong (Indian Health Service). The technique exists because the forgetting figures above are so bad, and because the nod is worthless as evidence.

At work you cannot say "teach that back to me," but you can get the same reading with a question that makes the playback natural. The ones that work put the burden on you, not on the listener's comprehension:

• "I have thrown a lot at you. What is the headline you are taking from this?"
• "Before we move on, can you tell me what you will do first, so I know I have explained it properly?"
• "Say back to me what you think I am asking for, because I am not sure I put it well."
• "What would you tell your team this meeting decided?"

What comes back is the only true score of how you communicated. If the playback matches your Habit 1 sentence, you are done. If it is close but missing the deadline, or has the wrong owner, you have found the specific thing to re-say, and you say only that. If it is wildly off, that is not the listener's failure; it is information about where you started.

Two warnings. Do not check the playback on everything; it is for the exchanges where a wrong playback costs something. And never correct a playback with "no, what I said was", because that converts a measurement into an accusation. "Right, and the piece I did not land is the date" keeps the burden where it belongs.

Habit 4: listen like a trampoline, not a sponge

Communication fails in the other direction too, and the common picture of good listening makes it worse. In 2016, Jack Zenger and Joseph Folkman analysed 360-degree assessments of 3,492 managers and compared the people rated in the top 5% as listeners against everyone else. The best listeners were not the quiet, nodding ones. They periodically asked questions that pushed the speaker's thinking forward, they made the conversation feel safe enough that disagreement could be aired, and they offered suggestions, which worked because they had earned the right to by then. The authors' image is that average listeners are sponges, absorbing, while great listeners are trampolines: the speaker's idea comes back with more height (Harvard Business Review, 2016).

For a communicator the point is this: the question you ask is also the playback. "So the risk is the vendor, not the timeline, is that right?" tells the speaker exactly what you took from them, gives them a chance to correct it, and moves the conversation one step. Silence tells them nothing. Zenger and Folkman also found that poor listeners were listening competitively, using the other person's turn to prepare their rebuttal. If you notice yourself composing while they talk, you have left the conversation, and whatever you say next will be a reply to something they did not quite say.

Habit 5: match the channel to the message

The Kruger experiments give a clean rule. Content survives text. Tone does not. Anything whose meaning depends on how it is said, feedback, disagreement, an apology, a request that could read as a reprimand, enthusiasm that could read as sarcasm, belongs on a voice channel: a call, a video call, a walk to someone's desk. The ten seconds of awkwardness in choosing the voice channel is cheaper than the two days of a colleague quietly reading your message as cold.

The reverse rule also holds. Anything the listener needs to retain precisely, a date, an amount, a sequence of steps, needs a text follow-up even when you said it out loud, because the forgetting figures apply to spoken words and not to the line in the email they can reread. The better communicators say the message by voice and then send the one-line version, which is, usefully, the Habit 1 sentence.

What does a better-communicated message sound like?

Here is a real shape of update, first as most people give it, then rebuilt on the five habits. The situation: a weekly check-in with your manager, and a deliverable is slipping.

Before. "So, on the onboarding project, we have been making decent progress, the engineering side is mostly there, there were a couple of things with the API that took longer than we thought, which, you know, is kind of typical, and design has been quite busy with the other launch so we have been waiting on a few of the screens, and I think realistically we are probably looking at, it might be a little tight for the fifteenth, but we are pushing on it and I will keep you posted. Also I was wondering if it would be possible at some point to get a bit more of design's time."

Count what the manager can repeat. Something about onboarding, something about design being busy, "a little tight," and a vague wish. There is no ask with a subject or a date. There is a hint. The real message, that the date is going to slip unless design time is reallocated this week, is present only as a feeling.

After. "The onboarding launch will miss the fifteenth unless design gives us two days this week. Engineering is done except for the three screens we are waiting on. The ask: can you get us two of Priya's days before Friday? If not, the honest new date is the twenty-second, and I would rather tell the client now than in ten days. Which of those do you want me to do?"

The playback sentence comes first and could be acted on alone. The ask has a person, an action and a deadline. The two outcomes are named so the manager knows what they are choosing between. And the closing question is the playback check in disguise: the answer will reveal immediately whether the manager understood the trade-off. If they say "just keep pushing," you have learned they did not take the message, and you re-say only the part that did not land: "To be clear, pushing without design time gets us the twenty-second, not the fifteenth."

Notice what was not needed: a bigger vocabulary, more confidence, better body language. The second version is shorter, plainer and, if anything, more polite, because it respects the manager's time and gives them a real decision. That is usually what "better communicator" turns out to mean in practice: fewer words, aimed.

How do you communicate better in the moments that matter?

The five habits apply everywhere, but a few recurring situations deserve their own note, because each one has a characteristic failure.

Meetings. The characteristic failure is contributing a topic instead of a message: "I just want to flag that there are some concerns around the vendor." Flag what, for whom to do what? Use the one-sentence rule hard here, because you have perhaps twenty seconds of the room's attention. "I think we should delay the vendor decision a week; I will have the comparison on Monday" is a contribution people can respond to. How to speak up in meetings has the timing and phrasing for getting that sentence into a crowded conversation.

One-to-ones and feedback. The characteristic failure is the hint, because feedback is uncomfortable and hinting feels kinder. "It might be worth thinking about how the report landed" is not feedback; the other person cannot repeat what you want changed. State the behaviour, the effect and the change you are asking for, then check the playback: "What are you taking from that?" Feedback is also the clearest case for Habit 5. Never put it in text first.

Disagreeing. The characteristic failure is listening competitively, assembling your case while they speak, and then answering a position they did not quite hold. Play back their position before you argue with it: "So your view is that the risk is cost, not timing." Half the time they correct you, and you have avoided arguing with a ghost. The other half they say "exactly," and they are now far more willing to hear your side, because they have evidence you heard theirs.

Remote and video. Everything leaks more. You lose the micro-signals that tell you a listener has drifted, and the muted grid gives you a nod from nobody. Playback checks become more important, not less, and the follow-up line in text is close to mandatory. How to be confident on video calls covers the delivery side, including the pause you need to leave for people to come off mute.

Being put on the spot. You will not have a prepared playback sentence, so make one in the pause. Take a silent beat, decide the one sentence, say it, and stop. The pause reads as thought, not hesitation, as how to use pauses when speaking explains, and the one sentence gives the questioner something to work with while you find the rest.

How do you practise becoming a better communicator?

You cannot practise this by reading about it, and you cannot practise it only in live conversations, because live conversations do not give you a transcript. The loop that works is the same one that improves any spoken skill: say it, capture it, score it against something specific, change one thing, repeat. The specific thing here is the playback.

The ten-minute playback drill. Pick a real message you have to deliver this week: an update, a request, a piece of feedback.

1. Write the playback sentence. One line, actionable alone.
2. Record yourself delivering the message as you would in the room, unscripted, 60 to 90 seconds.
3. Without replaying it, write down what you think a listener would repeat back.
4. Now get an actual playback. Ask a colleague to listen and tell you the headline, or hand the recording to a practice tool that summarises what it heard, or, at minimum, transcribe it and give the transcript to someone with no context.
5. Compare the playback with your sentence from step 1. Score it: matched, partly matched (right action, wrong owner or date), or missed.
6. Fix one thing, usually moving the sentence to the front or converting a hint to an ask, and record again.

Three rounds of this on one message takes ten minutes. Doing it on one message a week for a month changes how you build every message, because you start to hear the playback sentence while you are still planning the words. The drill also produces a number, the proportion of playbacks that matched, which is the only honest measure of whether you are getting better, since your own sense of clarity is precisely the thing the research says you cannot trust.

The weekly recording audit. Once a week, record one real meeting or call where you spoke for more than a minute, with permission, and listen back for the five habits. Count: messages that led with the playback sentence; hints that should have been asks; moments where you nodded or stayed silent when a question would have moved things; tone delivered on the wrong channel. You do not need to fix all of them. Pick the column with the biggest count and work on it for the next week. This is the same one-weakness-at-a-time loop that how to practice public speaking builds for presentations; the material here is just conversation.

What does a three-week plan to become a better communicator look like?

Week 1: find the sentence. Before every meeting or conversation that matters, write the playback sentence. Say it first. Do nothing else differently. By the end of the week you will have noticed how often you walked into exchanges with a topic and no message, and that noticing is most of the gain.

Week 2: convert the hints. Record two real conversations and count the hints. Rewrite each as an ask with a person, an action and a time, and use the rewritten form the next time the same request comes up. Add a playback check to one exchange per day, using one of the four phrasings above.

Week 3: fix the listening and the channel. In every one-to-one, play back the other person's position once before you respond to it. Move one thing you would normally have typed, a piece of feedback or a disagreement, onto a call, and follow it with a one-line text version. Run the ten-minute playback drill on the hardest message of the week.

After three weeks, re-record the same kind of update you recorded at the start and compare the playback match. The usual pattern is that the first-sentence habit arrives fast and sticks, hints take longer because they are tied to politeness reflexes, and the listening habit is the one that other people notice first, often before you feel you have changed anything.

Key takeaways

• A better communicator is measured by what the listener can repeat back, not by how clear the speaker felt. Those two things come apart more often than anyone expects.
• Speakers overestimate how well they are understood. In Savitsky and colleagues' 2011 study, spouses understood each other's ambiguous phrases no better than strangers did, and confidence did not predict accuracy.
• Decide the one sentence the listener must be able to repeat, and say it first. If the sentence cannot be acted on alone, it is a topic, not a message.
• Replace hints with asks: a named person, a named action, a named time. The closer the colleague, the more you hint, and the research says that closeness is not helping them decode you.
• Check the playback, not the nod. "What is the headline you are taking from this?" is the only real score of how you communicated.
• Listen like a trampoline: a question that plays back and advances their point beats silent nodding, per Zenger and Folkman's analysis of 3,492 managers.
• Tone does not survive text. Sarcasm was decoded at chance by email versus roughly 75% by voice in Kruger and colleagues' 2005 study. Put feedback and disagreement on a voice channel; put dates and numbers in writing.

Frequently asked questions

How do you become a better communicator at work?

Change your measure from "was I clear" to "what can they repeat." Before each meeting, update or request, write the one sentence the other person must be able to say back, and open with it. Turn every hint into an ask with a named person, action and deadline. After anything that matters, ask "what is the headline you are taking from this?" and re-say only the part that came back wrong. Put feedback and disagreement on a call, and follow spoken dates and numbers with a one-line message.

What makes someone a good communicator?

The listener walks away holding the message the speaker intended, and it cost them little effort to get there. Good communicators work backwards from that: they know the one sentence they need repeated, they say it first, they name the ask instead of hinting, and they check what arrived rather than trusting a nod. Clarity, concision and confidence are tools for producing that result, not the result itself. Research on miscommunication shows the speaker's own sense of clarity is the least reliable gauge available.

How can I improve my communication skills when speaking?

Practise with a transcript, because live conversation gives you no record. Record a real update or request, write the one sentence a listener should be able to repeat, then get an actual playback from a colleague or a practice tool and compare. Fix one thing, usually moving the key sentence to the front or converting a hint into an ask, and record again. Ten minutes on one message a week is enough to change how you build every message, and how to improve speaking skills covers the delivery mechanics underneath.

Is being a better communicator mostly about talking less?

Partly, but cutting words alone does not help if the remaining words are still a topic rather than a message. The rewrite in this guide is shorter than the original, yet the reason it works is that the first sentence can be acted on alone and the ask names a person and a date. Talk less, in the sense of removing the background narration that precedes the point; but what you remove matters less than what you put first.

Why do people misunderstand me even when I explain carefully?

Because careful explanation is judged from inside your own head, where the meaning is already known. Savitsky and colleagues found in 2011 that speakers expected their spouses to understand ambiguous phrases far better than strangers, and the spouses did not. Kessels' 2003 review found 40 to 80 percent of spoken medical advice is forgotten immediately. Careful is not the same as checked: ask the listener what they are taking away, and treat the answer as the real measure.

How long does it take to become a better communicator?

The first habit, leading with the sentence the listener must repeat, changes within a week once you start writing that sentence before each exchange. Converting hints into asks takes two to three weeks because it fights politeness reflexes. Listening with questions is the habit colleagues notice first, usually within a month. Measure progress with playback matches, not with how clear you felt, since feeling clear is exactly the thing the research says you cannot trust.

Conclusion

Being a better communicator is not a personality upgrade. It is a change of judge: from how clearly you felt you spoke to what the other person can say back. Decide that sentence, say it first, ask instead of hinting, check the playback, and listen with questions. Then record one real message this week and find out what actually arrived.