Stand Up Comedy Tips Filed 22 Nov 2026 · 16:40
How to Write a 20 Minute Comedy Set (Without Padding a Ten)
“In my experience most acts moving from five to ten peter out at about seven or eight minutes.” That’s Alfie Noakes, writing on the We Are Funny Project blog back in February 2022. It’s the most useful sentence in circuit advice, because it names the exact minute where the wheels come off, somewhere between seven and eight – the two minutes you were sure you already had. That’s the five-to-ten step he’s describing, though, not the ten-to-twenty one.
So here’s the whole argument of this piece, up front, before any of the structure: a 20 minute set gets assembled out of blocks that already work, and you count the blocks before you accept the booking. Everything below is about how to do that counting honestly.
How long is a 20 minute set, in chunks?
Waithe puts two to three chunks in a five-minute set, each one 90 seconds to two minutes long, and her route to 20 minutes is blending several of those five-minute sets together. That’s the number, and it comes out of two sentences in a GOLD Comedy guide written by the comedian Elsa Waithe in November 2017. Waithe says “Two to three CHUNKS are ideal for a five-minute set.” She also says “Each CHUNK should be 90 seconds to two minutes long.” Hold those together and the arithmetic does itself. At two minutes a chunk you need ten of them. At 90 seconds you need thirteen or so. Either way the answer is double figures, which is a very different shopping list from “four big bits”.
This is where most people’s planning falls down, in my experience of watching comics stare at a notes app in a venue car park. They think in bits, and they think they have more bits than they do. A bit that gets a laugh in the room where you wrote it is not a chunk you can bank. Her method is explicit: “As you progress, you can blend several five-minute sets together, and before you know it, you’ll have a 20-minute set!” The exclamation mark is hers, and the blending she describes happens “as you progress”, with no timescale attached to it.
Count yours tonight. Write every chunk you’d happily open with on a line of its own. Time each one with a stopwatch rather than a guess, then add them up. If the total is eleven minutes, you have an eleven minute act and a booking problem, rather than a twenty with ten minutes of improvisation wedged in the middle.
Four fives, stapled together
Cliff, writing on the Authentic Comedy newsletter in May 2024, puts the structural principle in one line. “A longer set should be a string of killer 5 minute sets.” That’s the thing to build: four fives you’d be happy to perform on their own, stapled together, with no arc and no callback-laden journey holding them up.
And the staples don’t need to be clever. Authentic Comedy’s answer to whether those blocks have to connect is flat: “No they don’t have to be related, or flow together.” Which will annoy anyone who has spent the last year reading about theme and throughline, but he’s explicit that the blocks don’t have to be related or flow together, and he warns against forcing material around a theme. (My own view, for what it’s worth: the forced theme is the single most common reason a decent ten becomes a bad twenty. You end up writing linking material, and linking material never gets a laugh.)
The same piece sets the entry requirement before any of this, and it’s higher than most acts want to hear. “Writing killer air-tight 5s should be your first goal”, it says, and step one is spelled out as: “Step 1: Create a 5 minute set that kills in every single room”. Every single room, in his words: a five that works everywhere, not just in a friendly one. If your tight five only survives friendly rooms, you’re still on brick one, whatever the running order in your phone says.
The old-new-old sandwich, per Ronny Chieng
This is the mistake I see most in a first 20: the comic puts the untested ten minutes at the front, while they’re still fresh and brave, and then spends the back half of the set apologising with their shoulders.
Ronny Chieng’s method, described in a Vulture interview and quoted in Matt Ruby’s Funny How newsletter in May 2022, has a name for the right shape:
It’s an old-new-old sandwich.
Proven, untested, proven. Funny How’s account of how Chieng actually deploys it is the operational bit: “Once new bits are ready, they get incorporated into a 10- to 20-minute club set, but with a cushion, Chieng says.” Two qualifications in one sentence, “once new bits are ready” and “with a cushion”, and between them they rule out what a nervous act actually does, which is open with the untested idea and hope.
Practically, for a 20, that means the first five and the last five are material you’d stake a paid booking on, and the new stuff lives in the middle third where a wobble is survivable. If a new chunk dies at minute nine you’ve still got eight minutes of proven set to climb out on. If it dies at minute two you’re doing crowd work for eighteen minutes, and crowd work is not a filler material strategy, it’s a separate skill you’re now performing badly under pressure.
The middle spot, the ‘spesh’ acts and the roof you shouldn’t blow off
British Comedy Guide’s guide to club bills, published in October 2018, describes the middle of a traditional club night as “a mix of newer and what used to be known as ‘spesh’ acts”. That’s the company you’re in, and that’s as far as the guide goes: it says who is on the middle of the bill, not what the slot is for.
There’s a wrinkle in that same piece worth knowing before you go out there swinging for the fences, because it describes what happens when the act before the headliner does too well: “that could prove a problem for the actual headliner, if the act before them blows the roof off the gaff.” Nobody will tell you this at the gig. Leave the roof on. As British Comedy Guide puts it, an act who blows the roof off the slot before the headliner could prove a problem for the headliner who follows them.
Where this sits on the ladder is the other reason to care. Noakes writes that “Once an act has something resembling a solid twenty minutes then they can, in broad strokes,” start operating differently, and Funny How describes the same escalator from the other end: “As the set grows, comedians start booking longer shows, first in clubs and then on tour in theaters.” The twenty is the length at which the conversation moves from spots to bookings, which is also the point you need to work out what to actually charge. I’m not giving you a UK fee figure here, because none of the sources above has one.
The three ways a first 20 goes wrong
Padding is the first. It’s Noakes’s seven-or-eight-minute wall, scaled up: an act with twelve minutes of real material takes a twenty and fills the gap with longer setups, repeated premises and the words “anyway”. The fix is arithmetic rather than courage. Ten to thirteen chunks, each timed, each already working in more than one room.
Talking faster is the second, and it’s the subtlest because it feels like energy. You’ve got fourteen minutes of material and twenty minutes to fill, so your body solves the problem by sprinting, which compresses the material further and gets you to minute eleven with nine to go and nothing in the tank. If you catch yourself doing this, the honest move is to drop the gig length rather than inflate the delivery.
Front-loading the untested is the third, and Chieng’s sandwich is the answer: the new chunk goes in the middle, with proven material either side of it. Worth saying that even a well-built 20 can die, in which case the getting-over-it part is its own skill and the material probably wasn’t the problem.
One more structural trap, which is the opposite error: deciding the twenty is a staging post to an hour and rushing it. Authentic Comedy is blunt about this. “I know an hour seems like a great achievement like running an marathon, but don’t rush to do it.” The typo is in the original, and the advice survives it.
How do you test a 20 minute set before the paid gig?
You test it live, in pieces, because video won’t do it for you. Waithe’s line about tapes is the relevant constraint: “Five minutes is also pretty much the maximum time a booker or producer will spend watching your tape.” So the twenty never gets proved on video; it gets proved in rooms.
What works instead is deeply unglamorous. Take your four fives out separately, to four different rooms, over several weeks, and notice which chunk dies when the room is older, or drunker, or sober and seated in rows. Then run the full twenty somewhere with no money attached before you run it somewhere with money attached. A Tuesday night in a room with eleven people in it is a better rehearsal for a December party gig than any amount of pacing around your kitchen, and it’s also how you find out which of your chunks only worked because the previous chunk warmed the room for it.
The December timing matters here, which is the other reason to have this sorted now. A Christmas party gig is the worst possible place to discover your 20 is a 14.
What to do with the list you just made
Count the chunks, time them, and put the number somewhere you’ll see it. If your proven five-minute blocks don’t add up to twenty yet, say what you’ve actually got when the offer comes. Noakes’s seven-or-eight-minute wall is just a measurement, and measurements are cheap compared with standing in a paying room at minute twelve holding nothing but the Gatwick parking bit you already know doesn’t work.