A great communicator clears four layers in order. Understand: did they follow it. Remember: can they recall it an hour later. Act: did they do something. Pass it on: can they retell it to someone who was not there. Above all four sits one question, what would make this best for the audience.
In a two-day session with a professional association in Washington DC in June, I asked a room of scientists and communicators to define a good communicator. The answers came fast: clear, confident, concise. All true. All completely unmeasurable. Nobody in that room could have told me whether they had been clear an hour earlier, because clarity as usually described is a feeling the speaker has, not a result the audience produces.
So here is the framework I use instead. Four layers, in a fixed order, and each layer has a test that somebody other than you gets to answer. That last part is what makes it useful. You cannot grade yourself on any of these.
What makes someone a good communicator?
Someone whose message clears all four of these, in sequence:
- Understand. Did they follow it?
- Remember. Can they recall it an hour later?
- Act. Did they do something?
- Pass it on. Can they retell it to someone else?
The order is not decorative. You cannot remember what you did not understand, you will not act on what you cannot remember, and you certainly cannot retell it. Each layer sits on the one below it, which means a failure at layer one guarantees failure at four. It also means that when something goes wrong, you diagnose downward rather than adding more effort at the top.
Layer one, understand: did they follow it?
Plain language does the work here, and plain language is not the same as simple language. It means the listener does not have to stop and translate. The moment someone stops to work out what a word meant, they are no longer listening to the sentence you are currently saying, and you have lost roughly ten seconds you will never get back.
The most common leak in technical organizations is not jargon. It is acronyms.
The usual advice is to ban them. That advice is wrong, and experts know it is wrong, which is why they ignore it. Acronyms are precise, they are how the work actually gets discussed, and stripping them out makes a specialist sound stilted and vaguely condescending. The problem was never the acronym. The problem is leaving one unexplained.
Say the acronym, then the three words that explain it, then keep moving. "The SLA, that is our service level agreement, has a 24 hour clause in it." No pause for permission, no apology, no detour. It costs about a second and a half. After the first use, the acronym is yours for the rest of the conversation.
What makes this work is that it does not condescend to the people who already knew. A definition delivered as a full sentence stops the room. A definition delivered as a comma barely registers, except to the person who needed it, for whom it is the difference between staying in the conversation and quietly checking out.
The test: did they follow it. The cheapest way to check is to ask one person to say it back in a single sentence. Do not coach them. Whatever comes back is what you actually communicated.
Layer two, remember: can they recall it an hour later?
Understanding is not memory. A room can follow every word you say and walk out holding nothing, and that failure is invisible to you because everybody nodded in the correct places.
Three things produce memory, and one habit multiplies all three.
Visuals give an idea a shape. That does not mean slides. A number written large, a gesture that shows a size, a physical object held up. Something the eye can attach to.
Analogies attach new information to something the listener already owns. This is why a good analogy feels like a shortcut: it is borrowing memory that is already built. The risk is choosing an analogy from your world rather than theirs, which lands as one more thing to decode.
Stories give a fact somewhere to live. A statistic on its own is a number. The same statistic inside thirty seconds of what happened to one person is a thing people repeat at dinner. There is more on how to build these in storytelling in presentations.
And then threes. Group your supporting points in threes. Not because three is magic, but because a list of three survives the trip home and a list of seven does not. Anyone who has taken notes in a meeting and then tried to summarize it in an elevator already knows this. If you have seven points, you have one point and six pieces of evidence, and your job is to work out which is which before you open your mouth.
The test: an hour later, unprompted, what comes back. Not what they can find in their notes. What comes back.
Layer three, act: did they do something?
This is the layer where good speakers most often stop, one step short of the point.
Most presentations end with "any questions." That is not an ask. It is a polite way to avoid making one. The audience has understood you and possibly remembered you, and now they file out with nothing specific to do, and by Thursday the whole thing has evaporated into a general good feeling about your competence.
An explicit call to action fixes it, and explicit means naming three things: who does it, what exactly it is, and by when. Vagueness here is not modesty. It is a transfer of work from you to them, and they will not do it.
That line has saved more calendar hours in my sessions than anything else I say. It is also a diagnostic you can run before you build anything: if you cannot name the action, you are not preparing a presentation, you are preparing a document you intend to read aloud. Send the document. I go into the mechanics of the close in how to end with a call to action.
The test: did they do something. Measurable, uncomfortable, and the only layer with a paper trail.
Layer four, pass it on: can they retell it to someone else?
Almost nobody designs for this layer, and it is the one that decides whether your message has any reach at all.
The person in the room is very often not the person who decides. Your message gets retold in a hallway, in a status update, in a summary email you never see, by somebody with none of your context and half of your motivation. Whatever survives that retelling is your actual message. Everything else was rehearsal.
What makes a message portable is structural simplicity. If your point only holds together with your slides attached, your delivery attached and your five minutes of setup attached, it is not going to travel. The thing that travels is short, ordered and repeatable: one message, three supports, one action. That is not a coincidence, it is the same architecture as layers two and three, which is why building for memory tends to build for portability at the same time.
The test: ask someone who was in the room to explain it to a colleague who was not. Then listen to what the colleague repeats back. That second-hand version is the only honest measure of your work, and the first time you run it, it will sting.
What sits above all four layers?
One question, asked before you build anything: what would make this best for them?
Not what would make this most complete. Not what would make me look most rigorous. Not what would cover me if somebody asks. Every one of those is a speaker centered question, and every one of them adds material that costs you at one of the four layers.
This holds whether the audience is a thousand people in a ballroom or one person across a desk. In 13 years producing at ABC, NBC and Fox, the segments that worked were never the ones where the guest had the most to say. They were the ones where the guest had worked out what the viewer needed. The full case for this is in audience centered communication, which is the foundation the four layers sit on.
How do you use the four layers this week?
Take the next thing you have to present, in any format, and run it against the four questions before you build a single slide.
- Understand: is there a single term in here I have not explained in three words?
- Remember: what is the visual, the analogy or the story, and are my points grouped in threes?
- Act: who does what, by when?
- Pass it on: could someone repeat this accurately without me in the room?
Most people find they fail at layer four and had assumed they were failing at layer one. That is the useful discovery, because it moves the work from adding more detail, which is what everyone instinctively does, to removing whatever is stopping the message from being carried. It also tends to make the presentation shorter, which nobody has ever complained about.
If your team needs to be understood, remembered, acted on and repeated accurately, that is exactly what a working session is for. Tell us what is coming up.
The four layers framework is Jess Todtfeld's, refined across 20 years of training and most recently in a two-day session with a federal agency in Washington DC on 13 and 14 August 2026. Examples in this article are generic and identify nobody.