A gambler's guide to giving talks

A bewildered audience is better than a bored one.

If you're good at something, they say, never do it for free.

Accordingly, nobody’s ever paid me to give a talk. But I’ve given enough free ones that I’ve realized that there are only two types of presentations: Forgettable ones, that waste away on YouTube with 18 views; and memorable ones, that bestow their creators with the lifelong benefits of having done something Good. 

Nearly every talk is the former, and nearly everyone goes to every talk expecting it to be the former. With the exception of presentations that are intentionally inflammatory or inappropriate, presentations have a floor for how bad they can be—bland and immediately forgotten—and people's expectations are an inch above that floor. 

That sounds bad, but it has a very useful implication: There's very little risk in taking chances. A daring talk that flops and a cautious talk that safely sticks to all the comfortable best practices are equally ignored. But stick the landing on something bold, and the upside is unbounded.¹ 

So, when giving a talk, for your sake and your audiences’, gamble. Here are some ideas—worth, probably, exactly what I was paid to write them down—for how to do it.

Be entertaining, not educational

People often say, at what they feel is some key moment of their presentation, “if you take anything away from this talk, this should be it.” To me, this sentiment is profoundly misguided. Most people don’t go to talks to learn,² most of the things that we have to teach aren’t that interesting, and the only thing that most people will take away from most talks is either a mild feeling of annoyance that it was so boring, or a pleasant feeling of surprise that it wasn’t.³ 

The most memorable presentations are the ones that acknowledge this reality. They’re the ones that are written as entertainment first, and that care about how the audience feels during the half-hour that they’re trapped in a conference hall with no discrete ways out rather than what the audience does after the leave. They’re the ones that worry more about keeping people’s attention than changing their minds. They’re movies, not documentaries.

Admittedly, “be memorable and entertaining” is can sound like a daunting task, and often is. But it is the task, and pretending otherwise only makes it harder. 

Go faster

Nearly every primer on public speaking reminds presenters to slow down. People talk quickly when they’re nervous, the guides say; you’ll know you’re going at the right pace when it feels way too slow.

This isn’t necessarily bad advice, but it’s often interpreted in a bad way. When people slow down their speech, they often slow everything down. They repeat themselves; they add pauses to let their points breathe; they write ten minutes of content for a twenty minute presentation, to force their pace to a crawl. 

If the point of a talk was to hammer one point home, this might make sense. But as entertainment, this tempo is excruciating. What’s more, if an audience realizes that they can stare at their phones for a few minutes and not miss anything important—it’ll get repeated later, and summarized on a slide at the end—they will. 

Instead, give me a talk that keeps me on my toes. Give me one that assumes I’m bored and distracted—and smart enough to keep up with the pace of a normal conversation. Give me a talk that is sixteen minutes long instead of twenty, and actually needs all sixteen of those minutes.⁴ Make me pay attention, because I’m afraid of what I’ll miss if I don’t.

The Kroto method

The most memorable talk I’ve ever seen was a lunchtime keynote by Harry Kroto, a Nobel laureate in chemistry, at an undergraduate research conference in Tallahassee in 2004. I have no idea why I was there, what the conference was about, or what he said—but I’ll never forget how he said it. I was in some giant ballroom, eating lunch with the other attendees at one of a couple dozen circular tables. When Kroto was introduced, we all half-glanced at him, expecting to sit politely while he bored us with a lecture about the importance of science, or something. 

Instead, he was mesmerizing. In thirty minutes, he machine-gunned through about four hundred slides, never breaking his brisk rhythm as he did it.⁵ 

It was similar to the Lessig method—lots of slides, each with a single word, short quote, or simple image—but with a crucial modification. Lessig’s slides tend to be inessential, used as a way to underline certain words as he says them. Kroto’s slides were more clever. They told a kind of parallel story, making indirect jokes or emphasizing his points with visual analogies. For example, had Lessig said, “AI has become a worldwide phenomenon,” he might show a globe, or the word “AI” in a giant letters. Had Kroto said the same thing, he might show a picture of a sold-out show on Taylor Swift’s Eras Tour, or made some callback to a previous mention of a different worldwide phenomenon he has referenced earlier.

This style had three effects. First, by using something like the Lessig method, he solved the riddle about pacing: The constantly changing slides made his presentation feel fast, even though he was talking at a relatively moderate clip. Second, his addition to the method—using pictures that told their own story—made it impossible to turn away because you didn’t know what you might miss. Blink, and the joke might be gone. And finally, the internal references and callbacks gave the talk a sense of dazzling cohesion, like an episode of Arrested Development that continually weaves a half-dozen plotlines together at once.

Some people might say that this is distracting, or that you should write talks that can stand alone without the slides. On the first point, I partially disagree. Though pairing this style with complex slides would be confusing, an engaged audience is plenty capable of following a talk’s main narrative and its visual footnotes at the same time. On the second point, I fully disagree. Use the medium—a talk and slides—to its fullest. Giving a presentation that can be heard without being seen is like creating a movie that doesn’t need the picture. It does nothing but make a hard job—keeping easily distracted people entertained—even harder. 

Rhyme off the beat

Early rap songs tended to rhyme on the beat, with simple rhyme schemes. This pattern creates a monotonous, sing-songy effect that follows a very predictable wave. At best, it’s dull; at worst, it’s clunky and forced.

Most talks have an analogous flow. Presenters change slides “on the beat”—they make a point, say they’re moving on to their next idea, and change the slide. Or worse, they put up a slide that’s meant to tell a joke, and then, with the slide up the whole time, deliver the setup and the punchline. 

Nothing about this is engaging. There’s no tension or suspense. It also gives the audience permission to tune out—if a slide's headline is uninteresting, you can safely drift away until the slide changes, and decide if you want to start paying attention again.

Better presentations rhyme off the beat. Slide transitions happen in the middle of sentences, without warning. More subtly, the best transitions have no awkward anticipatory hitch before changing slides—”and sooo…”—and no pleading pauses that give the audience time to “get the joke.” Like great rap songs, they expect their listeners to keep up.

This flow makes presentations more unpredictable and—especially if the talk uses a lot of slides—better keeps an audience’s attention, since they don’t know what they might miss if they look away.

About the jokes: Embrace obscurity

A riskier suggestion: Include jokes, and it’s fine if they’re obscure. 

Will everyone get the Eras Tour reference? Probably not, but the people who do get it will appreciate what feels like an inside joke. If a presentation has enough of these sorts of references—and with a lot of slides, there are a lot of opportunities to include them—most people will get at least a few. 

These kinds of Easter eggs are far more memorable than generic jokes that most people get. Everyone will understand Boromir saying, “One does not simply…deploy to production;” nobody will laugh at it. 

More importantly, to the audience, somewhat obscure references are a variable reward for paying attention. That’s much better for keeping people engaged than a steady dose of familiar memes.

Focus on the transitions

Most people seem to develop presentations in roughly three steps: First, they create an ordered outline of the points they want to make. Next, they turn each point into a slide, and fill the speaker notes with how they want to talk about it. And finally, they practice speaking to each slide. 

Instead, when people are creating presentations, they should spend more of their time working on the transitions between their points, and less on the points themselves. 

When you practice the points, it’s easy to turn a talk into a list of key takeaways, presented one after the other, with no narrative glue connecting them together. Because most presentations’ action items aren’t actually very novel or interesting,⁶ it’s very hard to keep an audience’s attention this way. Talks need a story to keep the audience engaged, and stories are told in the transitions.⁷ 

There’s also a practical benefit to rehearsing transitions: People are usually already comfortable talking about the key points in their presentations. Even if these were the most important parts, they’re often the easy parts. The unnatural thing about giving a talk isn’t riffing on some subject that you’re an expert in; it’s packing that subject into a cohesive chronology. We should practice that part.

Be careful with coaching—including this post

For huge talks—ten-thousand person keynotes, stadium speeches, televised addresses—presenters likely have to play to the medium.⁸ At such a distance, a speaker has to act authentic, not be authentic. Under such bright lights, wear makeup.

For most of us, however, our half-full classrooms and two-foot-high stages are intimate enough that the opposite is true: Unless you’re a professional dedicated to the craft, acting feels fake—and audiences are far better at sniffing this out than we are at covering it up. For this reason, I'm generally skeptical of any sort of prescriptive coaching about how to write and give presentations. When people go through a handful of generic coaching sessions, they often start to hide their personalities behind a few clumsy "best practices:" Speak slowly; follow the hero's journey; tell a humorous personal anecdote to connect with your audience. 

It’s safe advice—and devastatingly boring. If you're a CEO presenting financial results on an earnings call and more worried about getting sued than keeping people on the line, by all means, do these things. But if you’re trying to keep people from drifting off into the middle distance, nothing is worse than the mechanical affect of a sporadically coached speaker.⁹ Far better to just be yourself.

So—the best thing to do with this post may be to ignore it. While this style is what’s comfortable for me, it might be a bad fit for other people. 

But don’t ignore it because it’s risky; that gamble is exactly the point. Because most talks are performative busywork: The speaker gives it for themselves, either for their resume, their ego, or to hear themselves talk. The audience shows up out of a vague sense of obligation, and listens because they’re there. Everyone nods along, goes home feeling like an accomplished white-collar professional, and pretends that the whole charade is useful. 

Give us the red pill instead. Take a chance, show us down a rabbit hole, and find out how far it can take you.

1. Literally—the possible benefits include everything from making a few useful connections to becoming the president of the United States.

2. I’d estimate that, at any given talk, twenty percent of the attendees are there because they’re deeply interested in the subject at hand, and are hoping they might hear something unexpected. Another twenty percent of the attendees are trying to decide if they’re interested in the subject, and are using the talk to figure out if it’s worth learning more. Fifty-nine percent are there because their employer paid for their conference ticket; despite actually going for the free happy hours and three days away from the kids, they’d feel guilty for if they didn’t go to at least a few talks; plus, there’s a corporate policy that anyone who goes to a conference has to write a perfunctory book report on what they learned while they were there. And the final one percent is there like an eager college student, looking to absorb as much knowledge as they can, writing down every idea, excited to put them to practical use.

Regrettably, most presentations are written as though the entire audience is in the last group.

3. They might also take away a few pictures of slides that will rot in either their Photos app or in a Slack channel called #random-industry-chatter.

4. Nobody finishes a three-hour-and-29-minute movie that could be half as long; nobody stays engaged in an hour-long meeting that should’ve been fifteen minutes; nobody pays attention to a twenty-minute presentation that has six minutes of content.

5. This presentation isn’t the talk we saw, but it’s a similar style.

6. Takeaway one: We need to focus on business problems, not technology problems. Takeaway two: Technology alone won’t solve anything. Takeaway three: Except the technology I’m selling, which you should buy.

7. Moreover, working on transitions has a similar benefit as writing in prose: It tests ideas much better than bullets ever can. As Jeff Bezos says, compared to bulleted decks, full sentences force “better thought and better understanding of what’s more important than what.” 

8. Or this at least seems true? I have no idea though. The biggest crowd I’ve ever talked to was one that was mostly there because the conference happened to be serving lunch at the back of the tent that I was presenting in. But if you have experience with these sorts of talks, 1.) hi, hello, this is an obscure data blog, why are you here?, and 2.) is this actually true?

9. More precisely, I’d argue that the relationship between professional coaching and talk quality follows a long U-curve. Some coaching often makes a presenter worse, by sanitizing them of their own style. Eventually, with enough practice—far more than most of us will ever put in—the character that a presenter plays can become more interesting than the person that they actually are. But very few of us will get there, and are better off being ourselves than a half-drawn sketch of some cyborgian ideal.