A finished catalogue that kept pulling views
- 19videos
- 160Ksubscribers
- ~40Mlifetime views
- ~2.1Mviews per video, lifetime
Public counts for the channel, as of Sep 2026.
19 videos. That is the whole channel, and it has not published anything since the series ended. The channel is still public and still gaining views, and the 2 biggest videos on it hold over 22 million views and nearly 13 million.
Close to 40 million lifetime views sit on cuts of mine, on one channel, from a series that stopped uploading when it finished.
Views per video is the number that makes the point. On a soap channel with 16,000 uploads, an average is meaningless. Here there is nowhere to hide: 19 videos, and the average is around 2.1 million.
A catalogue this small has no filler in it. Every video is either found or it isn't, and the difference is almost entirely in how it is packaged.
A channel that started at nothing
Every other channel in the set already had an audience when I got to it. This one was new, and a new channel is a different job.
There is no subscriber base to publish into, so nothing has a floor under it. On an established channel a weak upload still finds a few thousand people who already follow it, and that cushion is doing more work than anybody notices until it is not there. Here every video started from zero and had to be found by strangers, while the series was still going out and the schedule did not move.
Building subscribers and views at the same time is the hard version of that, because they pull against each other. The videos most likely to be found by somebody who has never heard of the show are rarely the ones that persuade that person to subscribe, and a channel that only chases the first never builds anything it keeps.
I ran the Facebook page for the series alongside the channel, and used each to push the other. That is an unglamorous advantage and a real one: a new channel has no audience of its own to draw on, so an audience somewhere else is the only audience there is.
This is also the only one of them I had start to finish. I opened it, cut everything that went on it, ran it while the series was on air and left it in the state it is in now, and nobody has published to it since. Whatever that channel is today, it is the thing I built and the decisions I made, with nothing else layered on top.
For a channel that started at nothing and stopped when the series ended, it did well.
The edit and the channel, at the same time
-
Cutting the series
All 19 went out while the show was on air, which is the version of this job with a clock on it. A magic act has to read on a phone with none of the studio around it, and the reveal is a frame-accurate thing: land it late and the trick is a man standing in a room.
-
Running what it left behind
Titles and descriptions, playlists, how the catalogue answered a search box, and which of the 19 a stranger met first. When the series wrapped the uploads stopped and the same 19 videos were all there was.
The 2 videos that carried the channel
By their published titles.
Those 2 hold roughly 22.5 million and 12.9 million views (as of Sep 2026). Between them that is close to 9 in every 10 views the channel has ever taken, and the lifetime total across all 19 is a little under 40 million.
- 00:00The trick, already in progress. A magic clip has no time for an introduction.
- 00:06Establish the constraint: what makes this impossible.
- 00:15Withhold the method. Show the hands, never the explanation.
- 00:24The reveal, frame-accurate. Land it late and the trick is a man standing in a room.
- 00:30Hold on the reaction. It is the proof the trick worked.
A schematic of how a clip on this channel opens, not a reading off any particular video.
Running a channel with nothing left to upload
Every piece of YouTube advice assumes the next video. This channel did not have one.
A back catalogue is not a dead channel. It is a set of things people are still looking for, described the way things get described on the day they go up: in a hurry, to a broadcast schedule rather than to a search box.
So the work moves. With no upload slot to fill, the only levers left are the ones attached to what is already there: what each video is called, what the first line of its description says, which playlist it sits in, and which of the 19 a stranger meets first.
The 2 videos that pulled the most were not the ones the broadcast would have led on, and they are not the 2 I would have picked either. That is the useful thing about a frozen catalogue: it settles arguments. Nothing new is coming to muddy the numbers, so what performs is what genuinely travels rather than what happened to be promoted that week.
What a frozen catalogue teaches you about metadata
It is the cleanest test of packaging I have had. No new uploads, no promotion cycle, no format momentum: just a fixed set of videos and the words attached to them. Whatever moved, moved because of how it was found or because the video itself travelled.
That is the same question a long-form edit asks, one step earlier. The title and the thumbnail make a promise, and everything after has to keep it. This is the one channel where the catalogue was finished and I could no longer touch any of it, so the only variable left was the first. You find out very quickly how much of the outcome it was carrying.
The catalogue was finished and there was nothing left to recut. What was left to change was how they got found, and that turned out to be most of the job.