What Makes a Good Keynote Speaker: Four Signals

You have a slot to fill and a shortlist of names, and every one of those names has a showreel where the audience is laughing on cue. So what makes a good keynote speaker when the reels all look the same? Charisma is the usual answer, and charisma is mostly lighting and editing. The more honest test is what survives the minute after the talk ends, when people stand up, check their phones, and either repeat one sentence to the person next to them or say nothing at all.

I perform mentalism and I speak about perception and belief, so I watch both halves of this from the same stage. The difference between a talk that lands and a talk that evaporates is almost never the content.

It is usually four things, and each one can be checked before you sign anything.

What makes a good keynote speaker in the first few minutes?

Attention is not handed over because somebody read out an introduction. It is given because something early in the talk refuses to sit in the background.

Most speakers open by establishing credentials. The room has already decided whether to listen by then, and credentials are the slowest possible way to earn that decision. What works is a problem the audience already has, stated more precisely than they could state it themselves. Precision is the real applause line. When somebody describes your situation better than you do, you assume they know the way out.

In my case the opening tends to be a demonstration, because I can show a perception failure faster than I can explain one. The mechanism is not the mentalism, though. The mechanism is a question the room cannot answer, arriving before anyone has settled into politeness.

A speaker with no props can do the same with one sentence. Ask your shortlist how the talk opens. If the answer is "I introduce myself and run through the agenda", you have learned something useful for free.

How do I tell if the speaker will be believed, not just enjoyed?

Enjoyment is cheap. A room will laugh at almost anyone who is warm and well paced. Belief is the expensive part, and belief is what the event is actually buying, because nobody books a talk hoping the team behaves identically on Monday.

Belief has a texture you can feel. It shows up when the speaker says something slightly against their own interest. A consultant who tells you which of their services you do not need. A speaker who admits where the method fails. The audience reads that as evidence the rest is not a pitch.

So here is mine. If your agenda needs somebody to explain your industry, your product roadmap or your regulatory environment, I am the wrong booking. I work on how attention and certainty get built inside a human head, which travels across industries precisely because it belongs to none of them. If the slot needs domain depth, hire the domain person, and give me the one after lunch, where the domain person would struggle.

Does the framework matter, or is that just packaging?

Mostly packaging, I will admit. Many frameworks on speaker websites are a list of nouns arranged so the first letters spell a word, retrofitted to a talk that already existed.

The real test is whether the framework survives a question from the floor. A load-bearing structure lets the speaker answer something they did not prepare for, because the structure is how they think. A decorative one collapses and the speaker walks back to the script.

Mine is called WAND in English (in Hebrew it is קופס"א, which is a better acronym and an untranslatable one). It runs through attention, certainty, wonder, story and belief: the sequence a mind moves through when it decides something is true. When I work as a keynote speaker, the framework is not an illustration bolted onto the talk. It is what happened on stage in the first few minutes, slowed down and labelled so the room can see the seams.

Ask your shortlist to explain their framework without naming it. If they can, it is doing work.

How much does the slot change who I should book?

More than the topic does. The same talk is a different job depending on where it sits in the day.

• Opening slot at a conference. The room is cold, people are still arriving, and the job is to set the temperature for everyone who follows. Energy and a clear frame beat nuance.
• Mid morning, after two internal presentations. The room is full but saturated. The job is to change the texture: a different voice, a different rhythm, something that does not look like another deck.
• After dinner, or at the end of a year end event. People have eaten and some have been drinking. A talk here has to live closer to a show than to a lecture, or it loses the room.
• Team evening with no stage and a low ceiling. Often this is not a keynote at all. It is an entertainment booking wearing a lecture's clothes, and I would rather say so than deliver content to people who came out for a good night.

That last one matters. If what you want is the room laughing and still talking about it in the car park, a mentalist show is the honest product and nobody has to pretend afterwards that it was a lecture. The attention problem behind all of this is something I unpacked in a piece on why big conference halls lose the room so quickly, and the logic holds in a hotel ballroom just as well as it does in Las Vegas.

What should I actually ask on the call?

Four questions. You will learn more from them than from any showreel.

• "What is the one sentence you want people repeating afterwards?" If they need a paragraph, the talk has no spine.
• "What goes wrong in your talk, and when?" Anyone who has done this a lot has a failure mode. A speaker who claims none has either not worked enough rooms or is managing you.
• "What do you need from the room?" Sightlines, amplification, a lit audience rather than a blacked out one. Specific technical answers are a sign of stage hours. Vague ones are a sign of a studio.
• "Who in our audience will not like this?" Somebody never does. The honest answer tells you how they think about a room.

As for me, the only credential I lean on here is the boring one: I run these principles live, in front of paying audiences, as my working week. If the attention does not land, the magic does not work, and if the magic does not work, the show does not work. There is no way to bluff that in front of a room, which is a narrower claim than most speaker bios make and a more testable one.

What to check before you book

A short list to work through before anyone sends a contract.

• Write the one sentence you want people carrying out of the room, and see whether it needs a talk or a show.
• Decide where in the agenda the slot sits, and what happens immediately before and after it.
• Confirm amplification. It is recommended for any room beyond a small one, and it can be supplied if the venue cannot.
• Check the language the room works in. I present in Hebrew and in English only.
• Fill in the questionnaire I send before a lecture, about the organisation, its current challenges and what to emphasise. It is what the examples and demonstrations get adapted from.
• Choose the format that fits the day: a Zoom session, a shorter in person taste, a deeper session, or a series built one principle at a time.
• Hold the date early if the event is fixed. There is no minimum lead time, but a popular date goes to whoever books it first.

So what should you decide now?

Before you compare speakers, decide what the slot is for. Write that sentence down. If you cannot write it, no speaker can deliver it, and the briefing call becomes two people being polite at each other.

Then decide whether that sentence needs a talk or needs a show. Both are legitimate choices, and booking the wrong one is the most common expensive mistake I see. That decision is yours, not the speaker's.

When you have your sentence, send the date, the city or venue, roughly how many guests you expect, and where the talk sits in the agenda, through the contact form or the WhatsApp number on the lectures page. Enquiries go to Ori, my personal manager, who will go through the details with you and work out which of these is the right fit, including the possibility that the honest answer is somebody else.