The last post was about why callbacks work. This post is about how to actually build one. Because understanding why something works and being able to construct it yourself are very different skills, and I have the scars to prove the distance between them.
I spent three months trying to write my first effective callback. Three months. For something that, when it finally worked, took four seconds to deliver. That ratio tells you everything about comedy construction. The audience sees the tip. You build the iceberg.
Here is what I learned about the practical mechanics of writing callbacks, tested and refined through a lot of trial and error at corporate events across Austria. Some of this comes from studying Greg Dean’s work on joke construction, where he breaks down how setups create expectations and punches shatter them. Some comes from Ralphie May’s emphasis on tightening material — getting fewer words between the setup and the laugh. But most of it comes from doing it badly, figuring out what went wrong, and doing it slightly less badly the next time.
The Three-Part Structure: Plant, Reminder, Payoff
Every effective callback follows a three-part structure. You can get more complex than this — some comedians layer four or five callbacks on top of each other — but the fundamental architecture is always the same.
Part one is the plant. This is where you introduce the element that will later become the callback. The critical thing about the plant is that it must not feel like a plant. It has to be organic, natural, a normal part of whatever you are saying or doing at that moment. The moment you telegraph the plant — the moment you pause for emphasis, or say “remember this” with your body language even if you do not say it with words — you kill the callback before it is born.
Part two is the reminder. This is optional and depends on the length of your show. If you are doing a ten-minute set, you probably do not need a reminder because the audience’s memory is fresh. If you are doing forty-five minutes or an hour, a casual mid-show reference helps keep the element alive in the audience’s subconscious without making it seem deliberate. The reminder is the lightest touch of the three. It is a glancing reference, not a full revisitation.
Part three is the payoff. This is the callback itself — the moment where the planted element returns in a new context, with a new meaning, and the audience connects the dots. The payoff must be surprising. If the audience can predict exactly how the element will return, you have built a setup, not a callback. The payoff earns its laugh from the collision between the familiar element and its unfamiliar new context.
The Three-Mention Rule
Here is a rule I stumbled onto through experimentation that I later found echoed in virtually every comedy writing resource I studied: three is the magic number.
One mention is forgettable. The audience registers it but does not encode it deeply enough for a callback to land.
Two mentions create a pattern. The audience thinks, “Oh, that again.” This is the recognition phase, but without a twist, it is just repetition.
Three mentions complete the arc. The first establishes, the second reinforces, and the third subverts. By the third mention, the audience has enough context for the callback to carry maximum impact, but not so much that it feels overdone.
Four or more mentions risk becoming annoying. There is a fine line between a running gag and a crutch, and that line usually lives somewhere around the fourth repetition. Some performers can push past four — Mac King does — but they have decades of calibration behind them. For most of us, three is the right number.
The reason three works is deeply embedded in how humans process narratives. Stories have three acts. Fairy tales have three wishes, three brothers, three attempts. Jokes have three elements in the classic “a priest, a rabbi, and a…” structure. Our cognitive architecture is wired to find satisfaction in threes. Two feels incomplete. Four feels excessive. Three feels like a complete thought.
My Step-by-Step Process
Here is the actual process I now use to build callbacks into my set. I did not arrive at this process through theory. I arrived at it through failing repeatedly and noticing what eventually worked.
Step one: I write my set without any callbacks. This is important. The show has to work without the callbacks. If a callback is doing structural work — if the show does not make sense without it — then you have not built a callback, you have built a dependency. The callback should be a bonus, not a requirement.
Step two: I go through the set and look for what I call “orphan moments.” These are small details, phrases, reactions, or situations that arise naturally in the first third of the show and do not currently connect to anything later. A comment I make about a volunteer’s reaction. A phrase I use to describe an effect. A minor incident that happens during a routine. These orphan moments are callback candidates.
Step three: I ask myself, for each orphan moment, whether there is a place later in the show where that element could return with a shifted meaning. This is the key creative act. The plant is easy — it already exists as a natural moment. The payoff is the invention. Where can this element come back in a way that surprises?
Step four: I write the payoff line. Just the line. I do not worry yet about delivery, timing, or context. I write the words that will constitute the callback and ask myself one question: does this make me smile? Not the audience — me. If it does not make me smile at my desk in Vienna with no audience pressure, it is not going to make a room full of people laugh in the moment.
Step five: I test it. I put it in the show and see what happens. This is the only step that actually matters, because no amount of desk-work can predict what will land in front of a live audience. The first test is almost always rough. The callback either comes too early (the audience has not had enough time to forget the plant, so the return feels forced) or too late (the audience has genuinely forgotten the plant, so the callback lands flat). Calibrating the distance between plant and payoff is entirely a live-testing process.
Step six: I adjust. And adjust. And adjust again. A callback that gets a mild chuckle at the first test might get a huge laugh after I change the timing by five seconds, or after I change the phrasing by two words, or after I shift my physical position on stage. The difference between a callback that gets a polite nod and one that gets a room-shaking laugh is often microscopic.
What Makes a Good Plant
The hardest part of callback construction is not the payoff — it is the plant. The plant has to thread a very specific needle. It must be memorable enough that the audience retains it, but casual enough that the audience does not realize they are retaining it.
Here are the characteristics of a good plant, based on everything I have tested:
It is specific. Vague plants do not stick. “I mentioned something about a card” is not memorable. A specific detail — a particular word, a particular object, a particular reaction from a specific person — sticks in memory because specificity creates mental images, and mental images are easier to recall than abstract concepts.
It is emotionally charged. Even mildly. Plants that are attached to a moment of laughter, surprise, or minor embarrassment stick better than plants that are purely informational. If the audience felt something when the plant happened, they are more likely to remember it later.
It is not flagged. The moment you draw attention to the plant as a plant, you trigger the audience’s analytical brain. They start thinking, “Why is he emphasizing that?” And once they are thinking analytically, the callback cannot surprise them later.
It is early. The best plants happen in the first ten minutes of a show. This gives maximum distance between the plant and the payoff, which means maximum surprise when the element returns. Plants that happen too close to the payoff feel like setups, not callbacks.
The Mistake of Over-Engineering
Early in my callback experiments, I fell into a trap that I suspect many performers fall into: I tried to build an elaborate web of interconnected callbacks, like I was writing a mystery novel where everything connects at the end. I had five different elements planted in the first fifteen minutes, all designed to converge in the final routine.
It was a disaster. At a conference in Graz, I performed this over-engineered version and the audience’s reaction was confusion, not delight. They could not track five separate threads while also following the actual content of the show. The callbacks, instead of creating moments of connection, created moments of cognitive overload. People were trying to figure out what the callbacks meant instead of enjoying the show.
I stripped it back to one. One callback. That is it. One element planted early, returning once at the end. And it worked. The audience laughed, spontaneously, with the warm recognition that a good callback produces. One clean callback is worth more than five muddy ones.
Ralphie May was right about tightening material. The principle applies to callbacks as much as it applies to jokes: get the excess out of the way. Stop trying to be clever with layers. Be clean with one.
As I have gotten more experienced, I have occasionally used two callbacks in a show — one verbal, one physical — spaced apart so they do not compete for cognitive resources. But I have never gone back to the five-callback web. That level of complexity might work for a dedicated comedy show where the audience knows they are watching comedy and is prepared to track multiple threads. It does not work for a keynote with magic where the audience’s primary cognitive resources are allocated to the content.
Tying the Show Together
The ultimate purpose of a callback is not to get a laugh, though that is a welcome side effect. The ultimate purpose is to make a collection of separate pieces feel like a unified show. Without callbacks, a show is a playlist — a series of individual songs played one after another. With callbacks, it is an album — a designed experience where later tracks reference earlier tracks and the whole is greater than the sum of its parts.
This is why I keep working on callbacks even though they are difficult and time-consuming to construct. Because the difference between a set that is “a bunch of stuff Felix does” and a set that is “Felix’s show” lives in the connections between the pieces. And callbacks are the most powerful tool I have found for creating those connections.
When a callback lands, the room changes. It is not just entertained. It is unified. Everyone shares a reference, a memory, a moment. For those few seconds, they are not a collection of individuals watching a performance. They are a group, bound together by shared experience.
That is worth three months of work for four seconds of payoff. Every time.