Five videos account for roughly 70 percent of every view our channel has ever gotten, which is both encouraging and a little insulting to the other 300 we made. Looking back at our most viewed Slushy Bros videos, the winners share about four traits, and none of them are production budget. We pulled the analytics, argued about it in the garage, and wrote down what we actually think happened.
Some of these took eleven hours to shoot. One took forty seconds and was filmed by accident. The gap between effort and performance is the whole reason this post exists.
The Five That Carried Everything Else
Here is the rough shape of our top performers, in order, with the numbers rounded because exact view counts shift every time we refresh the page.
- The first-try shot (around 2.1 million views): a single attempt that landed, and we had the reaction on camera. You can watch the clip where he got it in on the first try if you have never seen the one that started the run.
- The friends don’t drink friends bit (around 1.4 million): three costumes, one bad decision, and a punchline that only works because of who is wearing what. It lives here.
- The driveway basketball crossover (around 890k): costumes plus a real sport, which is a combination we now lean on constantly over on our sports videos.
- The machine meltdown (around 640k): our slushy machine failed on camera and we kept rolling for four extra minutes.
- The silent elevator bit (around 510k): zero dialogue, which turned out to matter more than we expected.
None of these were our favorite videos to make. Two of them we almost did not post.
The First 1.5 Seconds Did More Work Than the Punchline
On every one of those five, the costume is fully in frame before the first second ends. On our underperformers, we spent two or three seconds setting up a situation before anything visually strange happened, and the retention graph falls off a cliff in exactly that window.
Our average 30-day retention on the top five sits between 82 and 91 percent at the three-second mark. The bottom quartile of our catalog averages closer to 54 percent at the same point. That single gap explains more variance than script quality, lighting or editing pace combined.
We wrote a longer breakdown of this in why people swipe away from your shorts, but the short version is that a viewer decides whether you are worth 20 seconds before you have finished your first word. Lead with the weird visual. Explain later, or never.
Sound-Off Comprehension Is the Quiet Variable
Our silent elevator video has no dialogue, no caption burn-in, and no music cue that carries the joke. It still outperformed six scripted bits we considered funnier, and we think the reason is that it reads perfectly with the volume off.
A large share of mobile viewing happens muted, especially on the TikTok side of our audience. If your joke needs audio, you are asking a chunk of viewers to opt in before they know whether it is worth it, and most of them will not.
Three things we check now before posting anything:
- Does the bit still land if you mute it and watch at half speed?
- Is there a physical action in the first two seconds, not just a facial expression?
- Can a stranger tell who the characters are without reading a caption?
Repeat Characters Beat Clever One-Offs
Four of our five best videos feature the same color lineup in the same roles. The one-off experiments, the ones where we tried a new premise with new rules, consistently landed in the 20k to 60k range no matter how good the writing was.
Related reading: The Difference Between a Funny Idea and a Viral Video Idea.
Related reading: How We Turn Inside Jokes Into Slushy Bros Videos.
For a closer look at this topic, see our guide: What We Learned From Getting Our First Million Views.
People are not subscribing to jokes, they are subscribing to characters they already have opinions about. Once a viewer has decided that Red is a liability and Yellow is suffering, every new video starts with free context we did not have to earn in the first two seconds.
That is also why our free merch drop every 10K subs works better than a generic giveaway. The thing people want is a specific character’s color, not a logo on a shirt.
The Machine Failing Outperformed the Machine Working
We planned a clean demo video about layering flavors. The compressor stalled, the mix came out as soup, and we filmed ourselves reacting to it instead. That accidental version has more than eight times the views of the polished attempt we posted two weeks later.
Real failure on camera is extremely hard to fake, and audiences are better at detecting staged chaos than most creators assume. The comments on that video are almost entirely people describing their own equipment disasters, which is the sort of thread an algorithm notices.
If you are weighing a countertop unit for the house after watching that mess, we wrote an honest take on whether a slushy machine is worth buying, including the maintenance nobody mentions.
Why We Keep Getting Mistaken for Slushy Noobz
A meaningful slice of our search traffic is people looking for Slushy Noobz, the duo built around Hamzah and Martin, and landing on us instead. We are not them. We get the mixup constantly, and honestly the name overlap has sent us a few thousand views we did not earn.
Slushy Noobz built their following on reaction and funny moments content, and queries like slushy noobz first clip, first clip Hamzah Martin and Hamzah Martin reaction show up in our search terms report every month. The spelling variant slushynoobz appears almost as often as the spaced version, which tells you how people actually type on phones.
Two lessons we took from that, and one of them stings:
- Name collisions are real distribution. If a bigger channel shares a word with you, some of their audience will find you. You still have to give them a reason to stay past three seconds.
- Borrowed traffic converts badly. Viewers arriving from a slushynoobz search subscribe at roughly a third of the rate of viewers who found us through our own content, because they wanted a different video entirely.
We do not chase those searches. Building a look people recognize in one frame has been worth more than any keyword overlap, which is the same argument we make across our Instagram presence where the thumbnail does even more of the work.
What We Changed After Reading the Data
The pattern across our best performers is not subtle once you stop defending your favorites. We now build almost every short around four constraints.
- Costumes visible by frame 30. No cold opens on an empty driveway.
- One joke per video. If there is a second idea, it becomes its own post instead of a second act.
- Muted-readable action. The physical beat carries it, audio sweetens it.
- Something can go wrong. Real stakes, even small ones, keep people to the end.
The videos that break out for us are rarely the ambitious ones. They are the short, legible, slightly stupid moments where a viewer can tell within a second what they are watching and why it might make your day marginally better.
Frequently Asked Questions
What Was the First YouTube Video?
The first YouTube video was “Me at the zoo”, uploaded on April 23, 2005 by co-founder Jawed Karim. It runs 19 seconds, was filmed at the San Diego Zoo, and has since passed 350 million views. It is also a decent argument that production value was never the point.
Who Are the Two Guys in Slushy Noobz?
Slushy Noobz is the duo of Hamzah and Martin, who are separate creators with no connection to Slushy Bros. Their channel leans on reaction clips and funny moments content rather than costumed sketch bits. People confuse the two names often enough that we see it weekly in our search data.
What Are the Top Most Watched Videos on YouTube?
“Baby Shark Dance” sits at the top with more than 16 billion views, followed by “Despacito” at roughly 8.8 billion, according to the regularly updated list of most-viewed YouTube videos. Children’s content and global music releases dominate the top 20 because both travel across language barriers with almost no friction.
How Many Subscribers Does Slushy Noobz Have?
Subscriber counts for Slushy Noobz shift frequently, so the number on their channel page is the only figure worth quoting. Third-party trackers often lag by several days and round aggressively. The same applies to us, which is why our milestone merch drops are triggered by the live count rather than a screenshot.
Watch the Ones That Actually Worked
Every video mentioned here is still up, failures included, and the newest bits land first on our Frozen series. If you want the color you keep defending in the comments, the shop has it.
