Playbook: Lightning Talks

Lightning talks are one of my favorite event formats because they force clarity.

For the purposes of this article, I’m talking about a 20-minute session: 15 minutes of prepared remarks and 5 minutes of Q&A.

That is long enough for someone to say something meaningful, but short enough that they cannot wander forever. The constraint is the whole point. A lightning talk forces the speaker to decide what they actually want to say.

I would use this format when the speaker has one clear thing to teach, show, explain, or argue.

That could be a specific lesson from scaling a company, a tried-and-true playbook, a tactical operating process, a mistake they made, a decision they changed their mind about, or a specific trend they are seeing before everyone else.

The format is especially good when you want a lot of useful ideas in a short period of time. A block of three lightning talks can often be more valuable than one loose panel because each speaker is forced to get to the point.

It is also a good format when your audience is senior and busy. They do not need everything explained from zero. They want the sharp version.

The best lightning talks tend to be tactical, specific, and earned.

The speaker should have actually done the thing they are talking about, and be able to speak to it clearly and authentically.

Do not use a lightning talk when the topic needs a lot of context.

Some stories need a fireside. Some topics need a panel because there are multiple legitimate perspectives. Some speakers are better in conversation than prepared remarks.

The right format for the right topic.

A lightning talk is also not the right format if the speaker does not have a specific point they can (or want to) speak to. If their idea is “the future of AI” or “lessons from leadership,” that is probably too broad. Those are categories, not talks.

If the speaker cannot articulate what the audience will walk away with, the talk is not ready. Do not force the format just because it is convenient. A vague lightning talk is worse than no lightning talk.

The most important part of a lightning talk is the topic.

What is the one thing this person knows, believes, noticed, built, learned, or changed their mind about that the audience would benefit from hearing?

That is the talk. The idea should be specific enough that it creates curiosity and makes people want to attend when they hear the title.

Bad lightning talk topics sound like this:

“Scaling GTM”

“Building Great Teams”

“The Future of AI”

“Lessons in Leadership”

“Product-Led Growth”

Those are not talks, they are non-specific categories that one can speak to without actually saying anything of substance.

Good lightning talk topics sound more like this:

“How we doubled our candidate acceptance rate”

“Our first AI agent failed in production: here’s what we learned”

“The pricing strategy that boosted our gross margins by 10 points”

Those have a point of view. They imply a story and an outcome. They give the audience a reason to listen.

Speakers need a clear brief in order to be set up for success in giving a lightning talk.

I usually tell them something like:

You have 15 minutes of prepared remarks and 5 minutes of Q&A.

The goal is to make one clear point.

Assume the audience is smart, but does not know your internal context.

Do not spend more than 30 seconds introducing yourself or your company.

Do not pitch the product, unless directly relevant to your topic.

Use real examples, and anonymize/abstract out any sensitive data or information.

End with one clear takeaway.

No speaker wants to be boring — if their talk ends up being boring, it just means they have not been told what the format requires for success.

The rough structure can be simple:

1–2 minutes: setup

4–5 minutes: the problem, belief, or observation

4–5 minutes: the example(s) or story

3–4 minutes: the lesson or implication

1 minute: the takeaway

5 minutes: Q&A

The exact structure can change, but the speaker needs a structure. The audience should always know where they are in the talk.

This is a hill I will die on.

Nobody needs three minutes on what the company does, your work history, or your family life.

If the audience needs context, or you want to slip in a personal joke or fun fact, give them one sentence and move on.

“We’re a Series B infrastructure company helping enterprises monitor AI agents in production.”

Or “I’m a CMO and mom of 3 so when I say this workflow gave me meaningful time back in my day, you know I’m not exaggerating.”

Done. Move on.

The audience is not there for a commercial. They are there for a useful idea. If the talk is good, people will look up the company later to learn more, or follow up with you directly. If the talk is bad, the company overview will not save it.

This matters even more with sponsor speakers.

A sponsor can absolutely give a great lightning talk. Some of the best speakers at events are sponsors. But the talk has to be useful on its own. If the company were not sponsoring, would you still want this person on stage?

If the answer is no, find another way to include them. But do not give them a lightning talk.

Specificity is what makes a lightning talk work. Generic advice is essentially worthless. I once watched a 30 minute presentation where the BIG takeaway was “as an executive, you need to communicate your vision to the company.” 😴

If the talk is about hiring strategy, tell us the mistake, the signal, the interview loops, what changed, and what the results were.

If the talk is about product, show the before and after. What did the old flow look like? What did you change? What happened? What did you learn?

If the talk is about AI, do not just say “AI is changing workflows.” Tell us which workflow, what broke, what worked, what humans still had to do, and what surprised you.

The audience should feel like they are getting something they could not get from scrolling X.

That does not mean every talk has to reveal confidential information, but the more specific the examples are, the more useful the lesson becomes.

Slides should support the idea, not carry the talk.

The best slides are simple: one sentence, one chart, one screenshot, one before-and-after, one visual, one memorable phrase.

The worst slides are dense. Small text. Multiple charts. Five bullets. Screenshots nobody can read. Anything that makes the audience decide whether to read or listen.

Or worse, reading the slides while listening to you read them aloud.

For a 15-minute talk, I would usually aim for roughly 8–12 slides. Fewer is fine. More can work if they are visual and fast-moving.

The organizer should review the deck in advance, and the speaker should practice the talk multiple times.

Not to make it pretty. To make sure the point is clear, the examples are strong, the slides are readable, and the talk actually fits the time you have been allocated.

A lightning talk with a bad deck is usually a bad talk.

The first minute matters.

Do not start with a long bio. Do not start with “I’m so excited to be here.” Do not start with “A little bit about us.”

Start with the idea.

“Six months ago, I thought our onboarding was working. I was wrong.”

“Our highest-converting sales deck was creating our worst customers.”

“The first AI agent we shipped technically worked. It still failed.”

“We stopped asking customers what they wanted. Our roadmap got better.”

That is how you get the room.

A lightning talk does not have time for a slow warmup. The audience decides if your talk is going to be good or bad within the first two minutes.

The first line should make the audience lean in.

For a lightning talk, Q&A is part of the product.

The 15 minutes of prepared remarks should create the questions. The 5 minutes of Q&A should sharpen the point.

That only works if the talk was specific. Vague talks create vague questions. Specific talks create better questions.

The moderator or host should have one or two questions ready in case the room is slow.

Good questions sound like:

“What did you try first that did not work?”

“What would you do differently if you were starting over?”

“What surprised you most?”

“Where does this break at larger scale?”

“What should someone in the room do if they want to try this next quarter?”

Do not let Q&A become a rambling audience monologue. Five minutes is short. Keep it tight.

If you have multiple lightning talks, the order matters.

Do not just order them by who confirmed first. Think about the arc/narrative.

Maybe you start with the broadest idea, then get more tactical. Maybe you start with the most provocative talk to wake people up. Maybe you end with the most practical or memorable one.

Do not put two nearly identical talks back to back unless the contrast is intentional. If two speakers are both talking about AI adoption inside enterprise teams, make sure they are making different points.

The session should feel like a curated playlist.

A few small things matter more than people think.

First, use a countdown timer. The speaker should know how much time they have left, and the host should know when to move them along.

Second, make sure the slides are tested. Fonts, videos, GIFs, embedded demos, weird aspect ratios — all of this can break. Test it before the room is full.

Third, use a confidence monitor if possible. It helps the speaker stay oriented without turning around constantly.

Fourth, make the session title and speaker information visible somewhere. People walk in late. People take photos. People forget names. Make it easy.

Fifth, if you are recording, get the audio right. Bad audio kills the usefulness of the content afterward.

None of this is complicated, but it is the difference between a session that feels professional and one that feels thrown together.

Lightning talks are very useful content if you plan for them.

They are short, focused, and usually built around one idea. That makes them much easier to turn into clips, recaps, quote cards, internal notes, newsletters, or follow-up posts.

But you need to decide that before the event.

If you are filming, tell the speaker. If sensitive topics are involved, align on what can be shared. Do not surprise someone later by posting a clip they did not know was being captured.

Afterward, thank the speaker with a specific note. Tell them what landed. If people came up afterward to ask about a point they made, tell them. Good speakers are worth building around.

The goal of a lightning talk is not to give someone stage time — the goal is to make one specific idea land.

That means choosing the right idea, pairing it with the right speaker, cutting the talk down to what matters, reviewing the deck, rehearsing for time, protecting the Q&A, and placing the session properly in the agenda.

A mediocre lightning talk is just a short presentation.

A great lightning talk is one idea the audience remembers all day.