The two of us filming — © 2026 Sumi Pak. All rights reserved.Between 9% and 17% of active YouTube channels ever cross 1,000 subscribers. Of the ones that do, only about a third ever reach the 4,000 watch hours monetization also requires.
We started @miguktv in 2020 and crossed 1,000 on July 27, 2026. Six years, a year and a half of which we didn't post at all.
Our most-watched video came out in early 2021. 23,000 views. Which means we published something twenty-three thousand people watched, and it still took five more years to reach a thousand subscribers.
This is what we had wrong in between.
1. First, can you trust your numbers at all?
The advice at the wall is always the same. Low click-through rate, fix the thumbnail. Low watch time, cut the intro. Both low, change your topic.
All fine — and all of it assumes you have enough data to read.
Five percent from 40 impressions and five percent from 4,000 impressions are not the same number. The first is just what happened that week. Channels at the wall are almost always looking at the first kind.
So the first decision isn't "what do I fix." It's "can I trust this yet?" Swapping thumbnails six times on a video with three-digit impressions isn't optimization; it's something to do with your hands.
Read comments instead. Videos with almost no views still get them. Too few to be data, but a view tells you someone clicked and a comment tells you why they came.
2. Split click-through rate by traffic source
| Where the view came from | Typical CTR |
|---|---|
| Search | 8–15% |
| Browse / home feed | 3–7% |
| Suggested videos | ~9% |
| Overall average | 4–5% |
A 4% from search is underperforming. A 4% from browse is fine. And on most channels search is only 15–40% of views — the rest is browse and suggested, so keyword work alone won't fix discovery.
3. The algorithm isn't what's stopping you
Things YouTube's growth team has said repeatedly:
- Videos are judged individually, not against your channel average.
- Old videos underperforming doesn't drag down new ones. No penalty box.
- Time off isn't penalized.
- Small channels still get tested aggressively when early signals are good.
We took eighteen months off and still got here. If you stopped for a few months and think you've killed your channel — you haven't.
4. When you copy your own hit, most people copy the wrong thing
This is where we lost the most time.
When a video works, you want to do it again. Almost everyone repeats the wrong element. They repeat the subject, when what actually worked was the position the video was sitting in.
Two things to check:
- Was anyone looking for this?
- Could only you tell it?
You need both. Plenty of demand but no exclusive claim means hundreds of people already made it and you're buried. An exclusive claim nobody's looking for means nobody watches.
Our biggest video came from something we'd been through ourselves, at a moment people wanted to know about it, when not many people could tell that story. Both conditions, by accident.
Then we repeated the subject. We made several more on the same theme — except those were things anyone could film, and plenty of people were filming them. Interest disappeared almost immediately.
We took the topic and left the conditions behind. That mistake cost us years.
5. It's easy to blame the wrong thing
Afterward we believed the problem was the audience: people who arrived for that one subject never watched anything else, and it broke the channel. I thought for a long time that this was my fault.
That belief turns out to be mostly wrong. YouTube has said that having a lot of subscribers who don't watch doesn't reduce your home-feed reach. There's a smaller real effect — uploads go to subscribers first, and if they ignore them your early signal is weaker — but that isn't a broken channel.
The channel didn't grow because what we made afterward was material nobody was looking for. Not because of who subscribed years earlier.
If you're currently blaming a mismatched audience from your past, look at what you published last month instead.
6. When you and the person you make it with want different things
Skip this if you work alone. If you don't, this is more often what kills a channel at the wall than anything algorithmic.
We're a married couple making one channel, and we wanted different things from the start. One of us wanted to make what people search for. One of us wanted to film the life we actually live. It sounds like a small difference. Pick up a camera and it produces two completely different videos.
We spent years alternating between the two approaches and it didn't work. Then we changed format entirely, stopped partway through, and didn't post anything for a year and a half.
During that gap we were traveling in Korea, and I cried hard in a hotel room, trying to decide whether we should split the channel and go our separate ways. We joke now about a digital divorce. At the time it wasn't a joke.
Where the resolution actually is
In that situation people try to settle who's right. That isn't the question.
We were each half right. Demand matters, and it has to be something only you can make. Those are exactly the two conditions from section 4. We spent years each holding one of them and arguing.
But I was sure of one thing: I would not have lasted making it his way. Not because it was boring — because it wasn't something I could keep producing year after year.
So I decided I was going with my own work, and I restarted the channel myself. I don't quit things easily. I've held onto my art that way for as long as I've made it, and I held onto the channel the same way.
A strategy that's correct and unsustainable loses to one that's mediocre and survivable, because only one is still running in year six. Almost all channel advice tells you what to make. Very little asks whether you can stand to keep making it. That's the question that decides who's still there.
So this thousand is a small victory, and I'll call it that.
And it wouldn't exist without Jinguk. This isn't my channel, it's ours. We built six years of it together and we fought about it together. We're making this the way we're making a life — together, until we die.
7. What actually moved the number
- July 11, 2026 — 779 subscribers
- July 12, 2026 — Episode 1 of Yoon Jongshin: New York Museum Tour
- July 27, 2026 — 1,000 subscribers
221 people in fifteen days. A fifth of six years' work arrived in the last two weeks.
The easy conclusion is that a famous guest did it. Half true. This series is the second time in six years both conditions from section 4 lined up — something people were looking for, and a story only we could tell. We're an artist couple who also work as tour guides, and that job taught us that nobody assembles New York's museums into one coherent thing for visitors. We were building videos around that when a friend who loves art turned out to be the right person to tell it.
The series brought people. Six years is why they subscribed. Nobody subscribes to a single video — they look at what's already there and decide whether you're worth following.
A collaboration can't create a channel. It can only make an existing one visible.
8. What's still ahead
1,000 subscribers is one of two requirements. The other is 4,000 watch hours in a rolling twelve months, or 10 million Shorts views in 90 days. A lower tier at 500 subscribers unlocks fan funding.
Since only about a third of channels make that jump, the harder part statistically starts now. A subscribe is one click; watch time only accumulates if people stay.
The short version
- Before fixing a metric, check whether it's big enough to mean anything. Otherwise read comments.
- Split click-through rate by traffic source.
- The algorithm isn't blocking you. Time off doesn't kill a channel.
- When copying your own hit, copy the conditions, not the subject. Was anyone looking, and could only you tell it?
- Don't blame an old mismatched audience. Look at what you made recently.
- If you make it with someone, the question isn't who's right. It's who can keep making it.
- Collaborations reveal a channel. They don't build one.
Episode 4 of the series drops Friday, July 31.
If you're somewhere in this right now — is it the numbers, the topic, or the person you're making it with? We sat at 779 for a long time, and before that we sat at nothing for eighteen months. Tell me in the comments.
If you enjoyed this, subscribe and leave a comment.
Comments
Post a Comment