What we checked
Between 2 and 5 October 2026 we read the public upload history of 66 YouTube channels run by people who teach what they do: realtors, financial advisors, therapists, immigration lawyers, tech and AI educators, fitness and nutrition coaches. They ranged from 10,200 to 771,000 subscribers, with a median of 72,050.
For each channel we pulled the last 25 uploads through the YouTube Data API, separated Shorts from long videos, and compared the channel's best long video from the last 90 days with its own median. Comparing a channel with itself matters: 20,000 views is a flop for one channel and a breakout for another.
| What we measured | Result |
|---|---|
| Channels checked | 66 |
| Subscribers | 10.2K to 771K (median 72K) |
| Typical long video | 19 minutes |
| Typical views per long video | 12,558 |
| Days between long uploads | 4 (median) |
| Posted no Shorts in their last 25 uploads | 21 of 66 (32%) |
| Posted 6 or more Shorts in their last 25 uploads | 33 of 66 (50%) |
| Best recent video at 2x their median or more | 50 of 66 (76%) |
| Best recent video at 5x their median or more | 21 of 66 (32%) |
| Best recent video vs. median, typical channel | 3.5x |
Your best Short is usually already filmed
A video that beat your median is a vote your audience already cast. The topic, the title and something in the first minutes held people better than usual. That is the cheapest place to look for a Short, because you are not guessing what might work. You are taking a moment that already did and showing it to viewers who have never seen the channel.
The channels in this sample had plenty to work with. The typical one put out a 19-minute video every four days, which is a lot of footage. Yet 21 of the 66 had not posted a single Short in their last 25 uploads. Their best moments only ever reached the people who were already subscribed.
How to find your breakout video in five minutes
- Open YouTube Studio, go to Content, and look at your last 25 long videos.
- Find the median view count: sort by views and take the one in the middle. Do not use the average, because one viral video drags it up and hides every other winner.
- Mark every video from the last 90 days with at least twice that number. Those are your candidates.
- Open the strongest candidate's retention graph. Look for the stretch where the line goes flat or bumps up, which means people kept watching or rewound. That stretch is your clip.
- Cut 30 to 60 seconds around one complete idea, starting on the claim itself, not the setup that leads to it.
What makes the cut work as a Short
- One complete idea. A Short that needs the rest of the video to make sense loses people at the moment it ends.
- Start on the payoff. Long videos earn the right to a slow start. A Short has about a second. Open on the sentence that makes someone stop scrolling.
- Captions, word by word. Many people watch Shorts with the sound off, and big captions also keep the eye moving.
- Keep the face in frame. Reframe the horizontal shot to 9:16 around the speaker, not around the center of the old frame.
- Stay under a minute to start. YouTube accepts Shorts up to three minutes since October 2024, but a tight cut is easier to watch to the end. After posting, the first number to check in Studio is how many people viewed the Short instead of swiping away.
What this data can and cannot tell you
This is 66 channels, found through our own searches for expert topics, all English-language and all active in the last few weeks. It is not a random sample of YouTube, and channels that grew past a million subscribers are not in it. It shows how often a breakout video exists and how often it goes unused. It does not prove that a Short will grow a channel. That depends on the cut, the topic and the audience, which is why we test with one Short first.
We will add channels to this study as we read them and update the numbers here, with the date of each update. Last update: 5 October 2026, as more channels were read.
Questions
Why compare with the median and not the average?
Because one viral video pulls the average up so far that nothing else looks good next to it. The median is the view count of your typical video, so anything at twice the median really did beat your normal.
How long should a Short cut from a long video be?
Long enough to hold one complete idea and no longer. For most talking videos that is 30 to 60 seconds. YouTube allows up to three minutes, but a shorter cut is easier to watch to the end.
What if none of my recent videos beat the median?
Look further back, to your top videos of the past year, or use the moment in any video where the retention graph rises. A question your viewers keep asking in the comments is also a good starting point.
Do I need new footage for Shorts?
No. Every channel in this study already had enough long footage to cut Shorts from. Cutting from a proven video is the cheapest test there is.
Sources
How this guide was madeDrafted with AI assistance at Cinemora. Every number comes from our own reading of public YouTube data through the YouTube Data API between 2 and 5 October 2026; nothing in this guide is estimated or borrowed from another study.