How the YouTube Shorts algorithm works
Shorts distribution runs on a simple loop — shown, then watched or swiped away. Most of what creators argue about sits outside that loop entirely.
8 minute read
It's a different system from long-form YouTube
The first thing to understand is that Shorts and regular YouTube videos are recommended by different machinery. Long-form YouTube is largely click-driven: a title and thumbnail earn the click, and watch time earns the next recommendation. The Shorts feed skips the click entirely — viewers don't choose your Short, it is placed in front of them. That removes titles and thumbnails from the equation almost completely and moves all of the weight onto what happens in the first seconds of playback. Advice that works for long-form packaging mostly doesn't transfer, which is why channels that are excellent at one format are often mediocre at the other.
The loop that decides everything: shown, then watched or swiped
Mechanically, Shorts distribution is a testing loop. Your clip is shown to a small batch of feed viewers, most of whom have never seen your channel. Each one either keeps watching or swipes away, and the ratio — along with how much of the clip people get through — decides whether the next, larger batch sees it. YouTube's own guidance to creators says essentially this: the system looks at whether people choose to keep watching versus swipe past. Everything else you can measure is downstream of that one decision, made in roughly the first second, over and over, by strangers.
Subscribers barely factor into reach
A Short from a channel with a hundred subscribers can outperform one from a channel with a million, because the feed tests content against non-subscribers by default. This is genuinely good news for small channels — you are not waiting in a queue behind bigger accounts — but it also explains a frustration that hits established channels: subscriber count does not entitle a Short to views. Every clip re-earns its distribution from zero. If your Shorts numbers swing wildly from post to post while your subscriber count stays flat, that isn't a malfunction; it's the design.
What retention means when a clip is 30 seconds long
In long-form, losing half your audience by the midpoint is normal. In Shorts, the equivalent failure happens in the first two seconds, and the bar is much higher: strong Shorts hold most viewers to the end, and the best ones get watched more than once because the loop back to the start is seamless. That is why the standard advice — lead with the payoff, cut the intro — is really retention advice. If your analytics show viewers leaving instantly, the problem is the opening frame and first line. If they watch most of it and leave just before the end, the ending is flabby. Those are different edits.
Reposts, watermarks, and recycled clips
Repurposing long-form into Shorts is fine — it's the standard workflow for podcasters and streamers, and plenty of large channels run on it. What platforms have pushed back on publicly is low-effort recycling that carries another app's branding: Instagram has said outright that it deprioritizes videos with TikTok watermarks, and YouTube's guidelines consistently favour content that doesn't look like it was uploaded by a bot from somewhere else. The practical rule: clip your own material, export it clean for each platform, and put real effort into captions and framing. The re-use itself is not the problem; looking recycled is.
How to actually use any of this
You cannot optimise the algorithm directly, but you can optimise the two moments it measures: the first second, and the decision to stay. Concretely — open on the strongest frame and line you have; cut everything before the point; caption it, because much of the feed watches on mute; keep it as short as the idea allows so finishing is easy; and judge each clip by its retention curve rather than its view count. Then publish enough of them that the testing loop has something to work with. A single Short tells you almost nothing; thirty tell you what your audience swipes past.
FAQ
Do hashtags help YouTube Shorts get views?
Not meaningfully. They can add minor context, but Shorts distribution is driven by watch and swipe behaviour in the feed. A clip that holds viewers needs no hashtags, and one that doesn't can't be saved by them.
Does posting time matter for Shorts?
Much less than on follower-driven platforms. Shorts are recommended continuously over days or weeks after upload, so a good clip posted at an odd hour still finds its audience. Consistency of output matters more than clock time.
Why did one of my Shorts get 100x the views of the others?
Because each Short is tested independently against non-subscribers, outcomes per clip are extremely high-variance. One clip passing more test batches than usual is normal, and it doesn't guarantee the next one inherits its reach.
Related pages
Keep reading
More Creator Growth →The best time to post on Instagram, according to four studies
Buffer says Thursday 9am. Later says 5am. Sprout says early afternoon. They analysed billions of posts and still disagree — here is why, and what to do about it.
8 minute readThe best time to post on TikTok, according to five studies
Buffer analysed 7.1 million posts and rates Saturday the best day. Sprout Social analysed 2 billion engagements and says avoid the weekend entirely. Both are right, for different accounts.
8 minute readThe best time to post on LinkedIn, according to five studies
Buffer's 4.8 million posts point to Wednesday at 4pm and rank Tuesday among the worst days. Sprout Social's 2 billion engagements call Tuesday the best day of the week. Here is why.
8 minute readTurn one long video into a week of posts
300 credits for 3 days · no card.
Start free