The difference between a funny idea and a viral video idea is simple to say and brutal to apply: a funny idea makes one person laugh, and a viral idea gives thousands of people a reason to send it to someone else. We have filmed plenty of bits that cracked up the whole backyard and then died at 400 views. The jokes were fine. The ideas were not built to travel.
Virality is a distribution problem wearing a comedy costume. Once we started separating the two, our hit rate stopped feeling like a coin flip.
A Funny Idea Needs Setup, A Viral Idea Explains Itself
Most funny ideas rely on context. There is a character you have to know, a running joke from three videos ago, or a punchline that only lands if you watched the first fifteen seconds with sound on.
A viral idea carries its own setup in the frame. Someone scrolling at 1am with the volume off should understand the premise before they decide whether to keep watching. If you need a caption to explain the bit, you have a funny idea and a weak video.
We test this by muting the first frame and asking a friend what they think is about to happen. When they guess right and still want to see it play out, we film. When they shrug, we rewrite the opening instead of rewriting the joke.
The Three Questions We Ask Before Filming
Every concept that makes it past the group chat has to survive three questions. None of them are about whether the idea is funny, because by that point we already think it is.
- Who sends this to someone? If we cannot name the exact type of person tagging a friend, the idea has no built-in distribution.
- What does the thumbnail moment look like? There should be one freeze-frame that looks strange enough to stop a thumb.
- Can it survive a second viewing? Rewatches are how short-form videos climb, so a joke that only works once is capped.
A concept that answers all three is usually worth the production time, and that time adds up more than people expect. We broke down how long it actually takes to make a 20-second comedy video, which is exactly why we filter hard before anyone picks up a camera.
People Share Identity, Not Punchlines
The videos that spread fastest for us were rarely the funniest ones. They were the ones that said something about the person sharing them.
A share is a small public statement. When someone sends a clip, they are saying “this is my friend group,” “this is my team,” or “this is exactly what my roommate does.” The laugh is the excuse, the identity is the engine.
That reframe changed how we write. Instead of asking what is funny about a situation, we ask which group of people would recognize themselves in it hard enough to tag three friends in the comments. A joke about one guy ruining a tailgate travels further than a joke about nothing in particular.
Visual Premises Travel Further Than Verbal Ones
Dialogue jokes are hostage to sound, captions and attention. Visual premises survive bad phone speakers, autoplay and people watching in class.
This is most of the reason costumes became part of our whole thing. A ridiculous outfit does the setup work in half a second, and nobody needs to be caught up on the lore to get it. We put together a list of funny group costumes that are not completely overdone for the same reason: the visual is the hook, and the bit rides on top of it.
If your idea reads as a sentence but not as a picture, it probably belongs in a podcast episode rather than a short.
Rewatch Value Beats The Perfect Punchline
Short-form platforms reward watch time relative to length, and loops count. According to YouTube’s own documentation on Shorts, replays are counted in views, which means a 12-second clip people watch twice is doing double duty.
A punchline ending kills the loop. The viewer gets resolution, feels satisfied and swipes away. An ending that cuts half a beat early, or that loops back into the opening frame, keeps people in the video without them deciding to stay.
We now edit the last frame before we edit the joke. Trimming 0.4 seconds off the end has moved average view duration more than any clever line we have written.
Comments Are A Different Currency Than Laughs
A funny video gets likes. A viral video gets arguments, corrections, inside jokes and people claiming their friend is the one in the clip.
You can design for that on purpose without being annoying about it. Leave one detail slightly wrong, make one choice that half the audience will disagree with, or set up a situation that people have strong opinions about already.
- A visible mistake that someone will feel compelled to point out
- A ranking or a call that splits the room, like naming the worst drink at a party
- An unresolved detail that makes people ask what happened next
- A recurring character people want to defend or roast
Comments pull a video back into circulation hours after it was posted, which is why we track them separately from likes.
How We Rebuild A Funny Idea Into A Viral One
Most of our best performing videos started as mediocre concepts that got rebuilt. The rewrite usually follows the same order.
- Move the weirdest moment to the first second. If the payoff is at second nine, start there and explain backwards.
- Strip the dialogue to under ten words. Anything left should work as an on-screen caption.
- Make the premise visible. Props, costumes, setting, scale. Something the eye reads instantly.
- Name the audience. Decide which specific group should feel seen, then write to them instead of to everyone.
- Cut the ending early. Leave the viewer a half-beat short so the loop does the work.
That process does not guarantee anything. It does raise the floor, so the bad videos still get 20,000 views instead of 400.
The Numbers We Watch In The First Two Hours
We stopped judging videos by view count in the first hour, because the view count mostly reflects the size of the initial test audience. The useful signals show up as ratios.
- Average view duration above 90 percent of the clip length usually means the loop is working
- Shares above roughly 1 percent of views suggests the idea has identity value, not only comedy value
- Comments above 5 per 1,000 views tells us the concept has something to argue about
- A swipe-away spike in the first 2 seconds means the hook failed, no matter how good the back half is
When two of those four land, we repeat the format rather than the exact joke. We wrote more about that call in our post on how to know when a video concept is worth repeating, because repeating the wrong thing burns an audience fast.
Why Platform Size Changes What Counts As Viral
Viral is relative to where you post. YouTube remains the most widely used online platform in the United States, with about 83 percent of adults saying they use it according to Pew Research Center, so a clip that spreads there reaches a broader mix of ages than the same clip on a single app.
We cut different versions for different homes, and the same premise gets a slower opening on one platform and a harder cold open on another. You can see how that plays out across our TV uploads compared to the shorter vertical stuff.
Frequently Asked Questions
How Long Should A Viral Short Be?
Most of our best performing shorts land between 11 and 22 seconds. Under 10 seconds rarely gives the premise room to breathe, and past about 30 seconds the completion rate drops off sharply unless there is a genuine payoff people are waiting for.
Can A Video Go Viral Without Being Funny?
Yes, and roughly half of what spreads has no joke in it at all. Satisfying processes, skill clips, arguments and nostalgia all generate shares, because the share is driven by emotion and identity rather than comedy specifically.
How Many Videos Does It Take Before One Hits?
In our first year it took about 30 to 40 uploads before anything cleared a million views. The useful part was not the hit itself, it was that by upload 35 we could tell within two hours whether something was working.
Does Posting Time Matter More Than The Idea?
The idea matters far more, though posting time can shift early performance by maybe 10 to 20 percent. A weak concept posted at the perfect hour still stalls, while a strong one tends to find its audience over several days.
Watch What We Test Next
We post the experiments that work and the ones that flop, usually within the same week. Come see the current run on our creators page and tell us which ideas you would have cut.



