'How long should Shorts be?' is the most-asked Shorts question and the most-answered badly. The honest answer: between 20 and 60 seconds for most content — but the real variable isn't the algorithm's preference, it's your hook's payoff timing.
This guide breaks down what retention data actually shows by length and genre, why the 3-minute Shorts format changed the math, and a testing method to find the length your specific audience completes.
What the retention curves actually show
Across thousands of Shorts, completion percentage is the dominant ranking signal, and completion is brutally length-dependent. A 25-second Short that 55% of viewers finish beats a 55-second Short that 30% finish — even if the longer one gets more total watch seconds.
The 2026 pattern: hooks under 2 seconds retain 15–25% more viewers than 4-second setups, and every additional 15 seconds of runtime cuts completion roughly 8–12% for educational and story content. Comedy and meme content decays slower because rewatching is the point.
Why 3-minute Shorts changed the game
YouTube now allows Shorts up to 3 minutes, and the algorithm treats completion percentage the same regardless of length. That means a 2-minute Short with 45% completion generates far more absolute watch time than a 30-second Short at 60% — and watch time is what Shorts surfaces broadly.
The catch: viewers swipe faster on longer Shorts unless the first 5 seconds prove the payoff is worth staying for. The 3-minute format rewards storytelling and punishes padding.
Length by genre: what works where
Quick wins (tips, facts, hacks): 20–35 seconds. One idea, one payoff, loop-friendly ending. Storytime and narrative: 45–90 seconds — long enough for a setup-payoff arc, short enough to finish. Tutorials and demos: 60–120 seconds, or split into a series where each Short ends on the next step.
Comedy and memes: 15–40 seconds with a loop point — the last frame should flow into the first so replays inflate completion.
How to find YOUR optimal length
Take your last 10 Shorts, bucket them by length band (under 30s, 30–60s, 60s+), and compare average completion and swipe-away points in YouTube Studio. Most channels discover one band clearly outperforms the others — then standardize your scripting to that band.
Aim your edits at the swipe-away point, not the average. If 40% of viewers leave at second 12, something specific happens at second 12 — a transition, a repeated phrase, a visual pause. Fix that moment before you change anything about length.
Tips
- Script backwards: write the payoff first, then the shortest possible setup that makes the payoff land.
- Use a loop ending on Shorts under 35 seconds — seamless replays count as completion and can push effective retention above 100%.
- Never pad to hit a length. A 22-second Short that fully delivers beats a 58-second Short that meanders, every time.
- Test one variable at a time: same format, same hook style, two length bands, four videos each — then commit to the winner.
Common mistakes
- Chasing the maximum length — Uploading 3-minute Shorts because the format exists, not because the story needs it — the swipe-away spike in the first 10 seconds poisons the video's distribution.
- Ignoring the swipe-away graph — Creators obsess over average view duration while the retention graph shows a mass exodus at one specific, fixable moment.
- One length for every format — A tips channel and a storytime channel have opposite optimal lengths. Copying another niche's format without copying their retention profile fails.
Example scenario
Illustrative example — not a real customer.
Situation: A fitness creator's Shorts average 48 seconds with 38% completion; swipe-away spikes at second 15 where a logo intro ends.
What changed: Cutting the intro and restructuring to a 32-second single-tip format lifted completion to 61% within four uploads — same content, same hooks, tighter payoff timing.
Frequently asked questions
Is there a minimum length for Shorts monetization?
Videos under 8 seconds aren't eligible for the Shorts revenue pool. Practically, aim for 15+ seconds — under that, there's rarely time for a hook and a payoff.
Do longer Shorts get less distribution?
Not directly — the algorithm scores completion and watch time, not duration. But longer Shorts naturally get lower completion, so the effective reach is usually lower unless the content genuinely holds attention.
Should I reupload long-form clips as 3-minute Shorts?
Only if the clip works standalone with its own hook. A segment that assumes the viewer watched the full video will lose the first 10 seconds of every viewer.