How Long Does a Short Drama Video Keep Getting Views? About a Month
We counted the views on 3,163 official uploads, waited 27 days, and counted again. A video in its first week gained 50%. One older than a year gained 0.4%, and a quarter of them gained nothing at all.
Almost everything on this site can be measured once. This cannot: to know whether a video is still being watched, you have to count its views, wait, and count again.
So we did. We read the view count on every upload we track, waited, and read them all a second time — a median gap of 27 days. Matching the two reads gives 3,163 videos with a before and after.
The answer is that a short drama video does almost all of its earning in its first month.
Growth in 27 days, by how old the video already was
| Age at the first read | Videos | Median growth | Median views added |
|---|---|---|---|
| Under 7 days | 188 | +50.2% | 1,331 |
| 7–30 days | 615 | +17.9% | 571 |
| 30–60 days | 572 | +7.8% | 237 |
| 60–120 days | 751 | +2.9% | 82 |
| 120–365 days | 750 | +1.3% | 81 |
| Over a year | 285 | +0.4% | 3 |
A video caught in its first week added half its views again over the following month. A video already older than a year added almost nothing: three views, at the median. Twenty-four per cent of the videos older than a year gained no views at all in 27 days.
The gap is 27 days for most of these videos and up to 41 for a few hundred of them, so we divided each video’s growth by its own gap before taking the median. That gives a daily rate:
| Age at the first read | Median growth per day |
|---|---|
| Under 7 days | 1.862% |
| 7–30 days | 0.666% |
| 30–60 days | 0.280% |
| 60–120 days | 0.103% |
| 120–365 days | 0.045% |
| Over a year | 0.013% |
A video in its first week grows at roughly 140 times the rate of one that is over a year old. The curve bends hard somewhere in the second month and is effectively flat by six.
What that means when you read a view count
An old video’s view count is final. If you are looking at something posted a year ago, the number under it is what it will always be, give or take a fraction of a percent. That is useful: it means the figure is stable enough to compare against other old videos.
A new video’s view count is not comparable to an old one’s. This is the trap. A clip posted last week with 3,000 views and a clip posted last year with 6,000 views are not two points on the same scale — the first is still climbing at nearly 2% a day and the second stopped moving months ago.
Nothing here tells you what happened inside the app. These are YouTube numbers, and YouTube is where these platforms advertise. A clip doing well as an advert says nothing about how many people went on to watch — or pay for — the series it came from. We have no visibility into the app side at all, and neither does anyone outside the companies.
Why we will not rank months against each other
This measurement also settles a methodological question that keeps coming up on this site, and it is worth being explicit about.
If you take a snapshot today and compare the median views of July’s uploads against September’s, July wins — not because July was better, but because July has had two extra months to accumulate. On the numbers above, a two-month head start is worth roughly a fifth of a video’s final total.
So: we do not publish month-versus-month view rankings, and when we describe a recent month we say the figure is a floor. The same caution applies to anything you read elsewhere that ranks “best performing months” from a single snapshot, unless it says how it handled the age difference. Our own attempt to measure whether short dramas are getting longer failed for a related reason: the sample changed shape over time, not the format.
What we measured, exactly
The first read is the view count we already held: 2,974 videos read on 1 September 2026, plus 255 read on 23 August and 156 on 18 August. The second read is today, 28 September 2026. We matched on video ID and kept every video with a usable count in both, which is 3,163 of them.
Three limits worth stating:
One window. This is a single 27-day interval, in one month of one year. It describes a decay shape, not a law.
These are the channels’ recent uploads. We hold up to 500 videos per official channel, weighted toward the newest, so the “over a year” group is the oldest tail of what is still on those channels rather than a random sample of old videos.
Medians, not totals. A handful of clips do enormously better than the rest, and totals are dominated by them. Every figure above is a median, which describes the typical video rather than the channel’s revenue.
The practical version
- If you are trying to judge whether something is popular, check how old it is first. Under a month, the count is still moving; over six months, it is finished.
- Do not compare a new clip’s views to an old clip’s. They are measured at different points on the same curve.
- A view count is an advertising number. It tells you the platform’s promotion worked on YouTube, not that the series is good — nobody here has watched these, and no platform publishes a rating we could borrow.
Both reads come from the official YouTube channels’ own public view counts, collected with the platform’s API. The figures in this piece use the 3,163 videos present in both reads.