A whole year publishing every week without a gap. In month five he reached three thousand views a video, and he was delighted. Month six: three thousand two hundred. Month seven: two thousand nine hundred. Month twelve: three thousand four hundred.
Seven months and the line was as straight as if somebody had drawn it with a ruler. So he changed the thumbnails, and nothing happened. He changed the style of his titles, and nothing happened. Then he started doubting the whole subject, and himself.
What was happening was not failure. It was a ceiling, and ceilings have reasons you can read.
A plateau means you succeeded in a narrow place
A flat number says one thing: you are serving the audience you reached completely, and you have not yet reached another one. The people who know you are watching, and the people who do not know you are not seeing you.
Which is why most attempts to break a ceiling fail: they improve the video for the audience already present, and that audience is satisfied already. The right question is not "how do I make a better video?" but "who is not seeing me, and why?"
Three kinds of plateau
| The kind | Its sign | What it means |
|---|---|---|
| A ceiling | Views flat, subscribers rising slightly | You have reached the limit of your current audience |
| The same audience | Views flat, share of new viewers low | The channel is closed around the people who know you |
| A slow decline | A small percentage down every month | Fatigue, or a shift in format or in subject |
Each kind has a different remedy, and the mistake is treating the second with the method for the first.
Four numbers before you change anything
Open your analytics on the last ten videos and write four numbers on a page:
- Impressions. Is your video being shown to people at all?
- Click-through rate. Do the people who see it click?
- The retention curve. Do the people who click stay?
- The share of new viewers. Is anybody reaching you from outside your circle?
Whichever number has flattened decides your work:
- Impressions flat and low: the problem is the subject itself, or a search nobody is making.
- Impressions rising and clicks flat: the problem is the title and the thumbnail.
- Clicks fine and retention weak: the problem is the first minute or the structure, and the reading is in The Retention Curve in YouTube Analytics: Where Viewers Leave and Why.
- Everything fine and subscribers flat: the problem is that the video gives no reason to come back, and the route is in How Do You Turn a Passing Viewer Into a Lasting Subscriber?
Anyone early on whose views were never high in the first place has a different case, with its four layers set out in Why Is Nobody Watching Your Videos?
Five real causes of a ceiling
You serve the same audience the same thing. Twenty videos answering the same question in different words. Whoever needed the answer has taken it, and whoever did not will not come. The fix is widening the subject into what borders it, rather than repeating it under new titles.
Your videos are closed rather than open. A video that answers a question and ends takes one view. A video that opens a new question takes two. That is a writing decision rather than an editing one.
Nothing turns a newcomer into a subscriber. The visitor watched and left, and nothing told him there are other things here like this one.
Every video is one format at one length. One format reaches one kind of person. Trying a second format is not inconsistency, it is looking for another door.
Your space really is small. Some subjects have a natural ceiling, and the number of people searching in them is limited. That is not your fault, but it does mean growth comes from expanding into a neighbouring subject rather than from more polish.
Ask your present audience about the absent one
The fastest route to the closed door is asking the people who came through the first one: what did you look for and not find? Who did you send my videos to, and why?
How you phrase the question matters, and the method is in How Do You Find Out What Your Audience Wants? One specific answer from a long-time viewer is worth a month of guessing.
A three-video experiment
Do not change the whole channel. Make three videos over six weeks:
- The first is safe: squarely what you do, so your audience stays with you.
- The second is adjacent: a subject for a nearby audience you do not serve yet, in the format you are good at.
- The third is a new format: your usual subject at a different length or in a different shape.
Then compare the three on one number only: the share of new viewers. That number is what tells you which door opened. Views alone will mislead you, because they may come from your existing audience.
The one-change rule
Change one thing, and judge after three videos rather than one. A single video can succeed or fail for an incidental reason; three reveal a pattern.
And if one of your videos does come in with unusual numbers, do not let it pass: the following seven days are what turn a wave into an audience, with the plan in The video that took off: what do you do in the seven days after?
Four things that never fix a plateau
- Deleting old videos. It raises nothing, and costs you views that arrive from search.
- Changing the channel's name and identity. It confuses the people who know you and does not bring the people who do not.
- Publishing more at the same quality. Doubling the count doubles the fatigue rather than the reach.
- Any service selling views or subscribers. It ruins all your measurements and harms your reach.
And tell yourself something true
Seven flat months after a year of work is not a sign that you are not cut out for this. Three thousand people watching something you made every week is a real number, even when it does not rise.
Do not open analytics hourly during those months either: repeated checking makes a plateau look like a decline and burns energy you need for the experiment. Once a week is enough, and once a month for a decision.
A ceiling is not a locked door. It is a door you have not yet looked for on the side you have not looked at.