The Creator GazetteThe business of creators

Data

Runtime, Cadence and Hit Rate: A Second Look at 335 Shorts From One Channel

Our first analysis of PopcornTheory looked at what the channel makes. This one looks at when, how often and how long. The 31-to-40-second short is the channel's engine, the top tenth of uploads carries 78 percent of all views, and busy weeks produce hits far more often than quiet ones.

By The Creator Gazette Staff5 min read

When The Creator Gazette published its analysis of the PopcornTheory numbers, the questions that came back were about mechanics rather than content. Does length matter? Does posting more often help or hurt? How much of a hit-driven channel is actually hits? We went back to the same dataset, all 335 shorts the channel published between March 8, 2025 and September 1, 2026, pulled again from YouTube's public Data API on September 25, 2026, and cut it three new ways.

A note on method before the numbers. Upload times are converted from YouTube's UTC timestamps to Brussels time, the channel's listed country being Belgium. Views are cumulative to the retrieval date, so recent uploads have had less time to accumulate than older ones. Nothing here is from the channel's private analytics; it is all public.

Length: the 31-to-40-second short does the work

PopcornTheory shorts run from 11 to 63 seconds. Sorting them into four runtime bands shows where the views come from.

RuntimeUploadsShare of uploadsViewsShare of viewsMedian viewsOver 1M
20 s or less288.4%27.5 million4.0%101,5205
21 to 30 s12437.0%236.7 million34.5%62,08722
31 to 40 s11734.9%346.1 million50.5%59,19431
41 s or more6619.7%75.3 million11.0%36,74115
Share of uploads and share of views, by runtime
Share of uploads and share of views by runtime bandGrouped bars: for each runtime band, its share of the channel's 335 uploads and its share of 686 million views. The 31 to 40 second band has 35 percent of uploads and 50 percent of views.15%30%45%60%Share of uploads, 20 s or less: 8%8%Share of views, 20 s or less: 4%4%20 s or lessShare of uploads, 21 to 30 s: 37%37%Share of views, 21 to 30 s: 35%35%21 to 30 sShare of uploads, 31 to 40 s: 35%35%Share of views, 31 to 40 s: 50%50%31 to 40 sShare of uploads, 41 s or more: 20%20%Share of views, 41 s or more: 11%11%41 s or moreShare of uploadsShare of viewspercent

Source: YouTube Data API, 335 PopcornTheory uploads, retrieved September 25, 2026.

The middle two bands hold 72 percent of the uploads and 85 percent of the views, but they are not equal. Shorts of 31 to 40 seconds are 35 percent of the catalogue and 50 percent of the views. Their hit rate, 31 of 117 passing a million views, or 27 percent, is the best of the four bands; the 21-to-30-second band manages 18 percent, the shortest band 18 percent and the longest 23 percent.

The longest band is the clearest underperformer. Shorts of 41 seconds or more are one in five uploads and one in nine views, with the lowest median in the table. The very shortest band has the highest median, 101,520, but it is also the smallest sample, 28 uploads, and it produced only five million-view videos.

Concentration: a tenth of the uploads, three quarters of the views

Hit-driven is an understatement. The top tenth of the catalogue, 33 shorts, accounts for 533 million of the channel's 686 million views, or 78 percent. The top three shorts alone account for 18.5 percent. The bottom half of the catalogue, 168 shorts, accounts for 0.7 percent.

The percentiles tell the same story from the other end. One short in ten has passed 7.6 million views. One in four has passed 523,000. The median short has 58,000. One in four has fewer than 29,000, and one in ten fewer than 19,000.

For a creator, the practical reading is that the typical upload is not the business. The business is the one upload in ten that travels, and the other nine are the cost of finding it.

Cadence: busy weeks find the hits

The channel's 335 shorts fall across 78 calendar weeks. Grouping the weeks by how many shorts went up in each shows a relationship between volume and outcome that is hard to argue with.

Uploads in the weekWeeksUploadsMedian views per uploadWeeks with a 1M+ short
1 to 2162146,4891 of 16 (6%)
3 to 42910154,06015 of 29 (52%)
5 to 6178937,9178 of 17 (47%)
7 or more1612488,42512 of 16 (75%)
Weeks that produced a million-view short, by how many shorts went up that week
Share of weeks with a million-view short, by uploads in the weekBars: weeks with 1 to 2 uploads produced a million-view short 6 percent of the time; weeks with 7 or more uploads, 75 percent of the time.25%50%75%100%Weeks with a 1M+ short, 1 to 2 uploads: 6%6%1 to 2 uploadsWeeks with a 1M+ short, 3 to 4 uploads: 52%52%3 to 4 uploadsWeeks with a 1M+ short, 5 to 6 uploads: 47%47%5 to 6 uploadsWeeks with a 1M+ short, 7 or more uploads: 75%75%7 or more uploadspercent of weeks

Source: YouTube Data API, 335 uploads across 78 calendar weeks (Europe/Brussels), retrieved September 25, 2026.

Weeks with one or two uploads almost never produced a million-view short: one week in sixteen. Weeks with seven or more uploads produced one three times out of four, and their median upload also did better, 88,425 views against 46,489. Part of this is arithmetic, since more tickets mean more chances, but the per-upload median rising with volume suggests the busy weeks were also the weeks the channel was in form.

Hits also cluster in sequence. Of the 73 shorts that passed a million views, the very next upload had a median of 266,657 views, and 29 of those 73 follow-ups, 40 percent, passed a million themselves. After the 207 shorts that finished under 100,000, the next upload had a median of 39,103 and only 27, or 13 percent, passed a million. Whatever the algorithm is doing, a hit is three times as likely to be followed by another hit as a miss is.

Timing: nights, and a Sunday effect that may be noise

More than half of all uploads, 188 of 335, went live between 22:00 and 04:00 Brussels time, which is mid-afternoon to late evening in the eastern United States. The pattern is consistent enough across the whole period to look like a rule rather than an accident.

By weekday, Tuesday and Saturday were the biggest days by total views, 148 million and 155 million, and Tuesday produced the most million-view shorts, 16 of 58 uploads. Sunday had the fewest uploads, 35, but the highest median, 139,382, more than double any other day.

Two cautions apply. Totals on a hit-driven channel are dominated by a handful of videos, so a single 47-million-view short moves a weekday's total more than fifty ordinary ones. And the weekday of an upload says nothing about the weekday of the views, which arrive over weeks. Treat the Sunday median as interesting rather than proven.

Engagement: longer shorts get more likes per view

Across the catalogue, viewers left 14.4 likes for every thousand views and 94 comments for every million. Likes per view rise with runtime: 13.0 per thousand for shorts of 20 seconds or less, 10.1 for 21 to 30 seconds, 16.5 for 31 to 40 seconds and 18.7 for 41 seconds or more. The band that travels furthest, 31 to 40 seconds, is not the one viewers like most per view; the longest band is. Length appears to trade reach for approval.

What it adds up to

Four findings survive the caveats. The channel's best format is a short of 31 to 40 seconds, which wins on views, on share of the catalogue and on hit rate at once. A tenth of the uploads carries nearly four fifths of the audience, so the channel's economics are the economics of the hit. Volume and hits move together: busy weeks produced million-view shorts twelve times more often than quiet weeks. And hits arrive in runs, with a follow-up to a hit three times as likely to hit again. None of this says what to make. It says, fairly precisely, how long to make it and how often.

Sources

Filed under: PopcornTheory, YouTube Shorts, Audience data

More from The Creator Gazette