How Long Should YouTube Shorts Be in 2026? We Analyzed 16,537 Videos

Should your YouTube Shorts be 15 seconds, 30 seconds, 60 seconds, or the full three minutes?

Ask ten creators and you'll probably get ten different answers.

Some insist that shorter is always better because viewers have less time to swipe away. Others argue that longer Shorts give you more time to tell a story, build curiosity, and deliver a satisfying payoff.

Both arguments sound reasonable.

But what does actual YouTube performance data show?

OverseerOS analyzed 16,537 short-form YouTube videos across 1,137 channels, comparing video duration with public view counts recorded approximately 7–30 days after publication.

The highest median view count appeared in the 16–30-second category: 36,970 views.

However, longer videos performed better than many creators might expect.

Videos lasting 121–180 seconds had a median of 31,847 views, placing them second among the seven duration categories we examined.

And when we compared different video lengths within the same channels, the advantage of shorter videos became considerably less convincing.

The answer isn't that every YouTube Short should be 20 seconds. It's that shorter videos can perform well, but length alone does not explain their success.

This study reveals what actually appeared in the data, why different Shorts-length studies reach conflicting conclusions, and how to choose a duration that fits your content instead of blindly following a formula.

Key Findings: The Best YouTube Shorts Length in 2026

Here are the main findings from our October 2026 research.

Source: OverseerOS original research, October 29, 2026. Public YouTube video and view-count observations.

An important qualification: this analysis uses OverseerOS's short-form classification, based on recorded video durations of up to 180 seconds. We did not independently verify the aspect ratio or Shorts-feed distribution of every video.

Therefore, the findings describe a selected sample of short-form YouTube uploads, not a complete census of videos confirmed to appear in the Shorts feed.

The direct answer: What is the best length for YouTube Shorts?

In our 16,537-video study, the 16–30-second range had the highest overall median views.

But the within-channel comparisons did not establish that 16–30 seconds consistently beats longer videos.

For a creator testing a new Shorts format, 16–30 seconds is a reasonable starting point for a single-idea video.

Use 30–60 seconds when the explanation needs more development.

Use 60–180 seconds when the story, demonstration, or educational payoff genuinely justifies the additional time.

Choose the shortest duration that delivers the complete viewer promise, not the shortest duration you can force the video into.

YouTube Shorts Length Comparison: All Seven Duration Categories

We grouped the 16,537 videos into seven duration ranges.

Then we calculated median public views and the share of videos that had reached at least 100,000 views by the observation checkpoint.

The percentages describe the observed sample. They are not the probability that a new video of a particular length will reach 100,000 views.

Three things immediately stand out

First, 16–30 seconds had the strongest median performance.

Second, the relationship between duration and views was not a simple downward slope.

If shorter were universally better, we would expect the shortest category to outperform every longer category.

That did not happen.

Third, the 121–180-second category performed surprisingly well in the pooled comparison.

Its median was considerably higher than the 61–90-second and 91–120-second categories.

That result deserves investigation, not an automatic conclusion that three-minute Shorts are better.

The channels and subjects producing longer videos may be different from those producing shorter ones.

An observed difference between duration categories does not establish that changing a video's length will cause its views to increase.

How We Analyzed 16,537 Short-Form YouTube Videos

The research was conducted on October 29, 2026, using public YouTube information collected by OverseerOS.

We wanted to avoid a common problem in Shorts research: comparing videos at wildly different ages.

A video published yesterday should not be directly compared with one that has been accumulating views for two years without accounting for that difference.

Step 1: Select short-form videos

We started with records classified as short-form in OverseerOS's video research database.

Each qualifying video needed:

• A valid YouTube video ID.
• A valid channel ID.
• A recorded duration between 1 and 180 seconds.
• A public view-count observation.
• A video age between 7 and 30 days at the observation.

This produced 16,537 qualifying videos from 1,137 distinct channels.

We retained videos with zero recorded views rather than excluding them automatically.

Step 2: Use one observation per video

Some videos had been checked multiple times.

Counting every observation separately would give frequently monitored videos more influence.

We therefore selected one observation per video.

Our preferred observation was the one closest to 21 days after publication.

When multiple observations were equally close, we selected the more recent observation.

The median video age in the final sample was 18 days.

This is more comparable than mixing recent uploads with lifetime performance, but it is not an exact day-21 cohort.

Step 3: Compare seven duration ranges

We divided videos into:

1. 1–15 seconds.
2. 16–30 seconds.
3. 31–45 seconds.
4. 46–60 seconds.
5. 61–90 seconds.
6. 91–120 seconds.
7. 121–180 seconds.

For every category, we calculated:

• Number of qualifying videos.
• Number of represented channels.
• Median public views.
• 75th-percentile public views.
• Number of videos reaching 10,000 views.
• Number reaching 100,000 views.
• Median video age.

Step 4: Reduce channel-composition bias

A large channel's videos can earn more views than a small channel's videos for reasons unrelated to duration.

To investigate that problem, we performed two additional analyses.

First, we calculated the median views for each duration category separately within each channel, then summarized those channel-level results.

Second, we compared duration categories within the same channels, requiring at least three qualifying videos in each category.

These checks were essential.

They made the final conclusion substantially more cautious than the pooled result alone would suggest.

Finding 1: The 16–30-Second Category Had the Highest Median Views

Among the seven duration categories, videos lasting 16–30 seconds had the highest median public view count.

36,970 views.

For comparison:

• 1–15 seconds: 21,913 views.
• 31–45 seconds: 22,947 views.
• 46–60 seconds: 20,938 views.

The 16–30-second category also had the highest observed share of videos above 100,000 views, at approximately 38.0%.

That makes it the leading category in this particular pooled sample.

Why might this length work?

A 16–30-second video can provide enough time to establish a recognizable question, demonstrate something interesting, and deliver a clear ending.

It is also short enough to avoid extensive setup.

For example, a 25-second science explainer might use:

0–3 seconds: Establish the question.

3–12 seconds: Introduce the surprising explanation.

12–22 seconds: Show the evidence or demonstration.

22–25 seconds: Deliver the conclusion.

That's an illustrative structure, not one directly tested by our dataset.

Another topic might require a completely different pacing strategy.

The point is that 16–30 seconds can accommodate a complete, focused idea.

Does this prove 20 seconds is the optimal length?

No.

We measured video durations and public views.

We did not randomly assign identical video ideas to different lengths.

We also did not observe private audience retention, swipe behavior, watch time, or completion rates.

The videos in different categories may differ in subject, production quality, audience size, and channel composition.

The strongest supported statement is:

In our selected short-form sample, the 16–30-second category recorded the highest median public views.

That is a useful benchmark.

It is not an algorithmic rule.

Finding 2: The Shortest Videos Did Not Have the Highest Median Views

A common assumption is that making a Short extremely short automatically improves its performance.

Our data did not support that as a universal rule.

The 1–15-second category contained:

• 2,138 videos.
• 323 channels.
• A median of 21,913 views.

That was substantially below the 36,970-view median of the 16–30-second category.

It was also below the median for videos lasting 121–180 seconds.

The practical problem with making everything too short

Some ideas can be communicated in ten seconds.

Others cannot.

Imagine a video explaining why a particular engineering structure works.

You may need to establish the problem, demonstrate the mechanism, and explain why it matters.

If you compress that into an arbitrary ten-second limit, viewers may not understand the answer.

A very short clip may create curiosity without delivering much value.

That does not mean short clips are bad.

It means duration should fit the subject.

When 15 seconds or less makes sense

Consider this range for videos centered around:

• One quick visual reveal.
• A simple demonstration.
• A single fact with an immediate payoff.
• A brief reaction.
• A fast transformation.
• One clearly defined before-and-after comparison.

For a topic requiring context, a longer duration may communicate the promise more effectively.

Our research cannot determine which creative approach is superior.

But it challenges the assumption that minimizing duration always maximizes views.

Finding 3: Two-to-Three-Minute Videos Were Surprisingly Competitive

One of the most interesting findings involved videos lasting 121–180 seconds.

This category contained:

1,751 videos from 348 channels.

Their median public view count was:

31,847.

That placed the category second in our pooled ranking.

It also had a 100,000-view share of approximately 33.8%.

This is important because many creators still think of Shorts as videos that must finish within one minute.

That is outdated.

According to YouTube's official three-minute Shorts documentation, eligible square or vertical uploads can run up to three minutes.

For standard channels, the broader three-minute eligibility rule took effect for uploads on or after October 15, 2024.

Why might longer short-form videos perform well?

Some subjects benefit from a more developed story.

Examples include:

• A miniature documentary.
• A historical event with a clear beginning and ending.
• A product demonstration.
• A scientific explanation.
• A multi-step transformation.
• A narrated personal experience.
• An educational story with a satisfying conclusion.

A 150-second video may give the creator enough time to develop the subject without stretching it into a conventional long-form upload.

But there is a critical limitation.

We did not confirm that all 1,751 videos appeared in the Shorts feed, nor did we control for differences between their channels and those publishing shorter videos.

Therefore, the result does not establish that extending a 30-second Short to three minutes will increase its views.

Longer Shorts also have different copyright considerations

YouTube's official guidance states that Shorts over one minute with an active Content ID claim can be blocked globally.

That is particularly important for creators using copyrighted music or third-party footage.

If you produce Shorts longer than 60 seconds, verify that your footage, music, and other assets are properly licensed for the intended use.

A longer runtime provides more storytelling room.

It does not remove copyright requirements.

Finding 4: The 16–30-Second Advantage Became Much Smaller Within the Same Channels

This was the most important methodological check.

The pooled data showed 16–30 seconds in first place.

But comparing different channels can exaggerate differences.

Perhaps channels that regularly publish 20-second videos are simply more successful than channels that publish 90-second videos.

To reduce that problem, we compared duration categories within the same channel.

Comparing 16–30 seconds with 31–60 seconds

We identified channels with:

• At least three qualifying videos lasting 16–30 seconds.
• At least three qualifying videos lasting 31–60 seconds.

That produced 183 comparable channels.

The shorter category had an advantage.

But it was much smaller than the pooled comparison might suggest.

A median channel-level ratio of 1.09x indicates a relatively modest difference in the typical qualifying paired-channel comparison.

It is not evidence of a massive, universal advantage.

Comparing 16–30 seconds with videos longer than one minute

We repeated the comparison using videos lasting 61–180 seconds.

This produced 85 channels with sufficient uploads in both groups.

The shorter category again had the advantage in slightly more than half the channels.

But plenty of channels performed better with longer short-form videos.

This is a major reason not to treat the pooled winner as a prescription.

What happens when we compare 16–30 seconds with 121–180 seconds specifically?

We found 26 channels with at least three qualifying videos in both categories.

The result was almost perfectly balanced.

13 channels favored 16–30 seconds.

13 channels favored 121–180 seconds.

The median ratio across the paired channel comparisons was approximately 1.09x.

But the equal 13–13 split makes it especially difficult to justify a universal claim that short videos outperform two-to-three-minute videos.

The sample of 26 channels is relatively small.

Still, it illustrates the importance of looking beyond pooled averages.

Different channels can succeed with very different runtimes.

Finding 5: Channel-Balanced Results Changed the Ranking

We also performed a separate channel-balanced analysis.

Rather than giving every video equal influence, we calculated each channel's median views for each duration category.

Then we summarized the channel-level medians, requiring at least three qualifying videos per channel-category combination.

The results were different from the original pooled ranking.

In this channel-balanced comparison, the 121–180-second category had the highest median of channel medians.

That is the opposite of the simple "shorter always wins" narrative.

However, this is not a perfect controlled comparison either.

The channels represented in each duration category are not identical.

The most reliable direct comparisons are the within-channel tests from the previous section.

What the channel-balanced analysis establishes is narrower:

The 16–30-second lead in the pooled sample was not robust to every reasonable way of weighting channels.

That is a meaningful result.

If you had only seen the pooled table, you might assume 20-second videos were clearly superior.

Looking deeper reveals a much less definitive picture.

Finding 6: The Leading Duration Was Similar Across Different Video-Age Windows

Even though we restricted observations to videos aged 7–30 days, the sample still included videos at different stages of their first month.

To test whether the results were unusually sensitive to observation age, we performed additional comparisons.

Videos observed around days 14–22

The 16–30-second category remained the pooled median-view leader.

We also examined younger and older portions of the 7–30-day observation window.

For the 16–30-second category:

• Videos observed at ages 7–14 days had a median of 34,356 views.
• Videos observed at ages 15–30 days had a median of 39,436 views.

The results show that the leading pooled duration was not simply an artifact of one category containing substantially older videos.

But these are different videos in each age subset.

They are not matched longitudinal measurements of the same uploads.

What this means

The pooled 16–30-second advantage was reasonably persistent under these sampling checks.

The within-channel advantage was more modest.

Both findings matter.

A responsible interpretation should acknowledge both instead of choosing whichever supports the strongest headline.

Finding 7: The Winner Changed Depending on What We Measured

Suppose you define success as:

"The duration category with the highest median public views."

In our primary sample, the winner is 16–30 seconds.

Now suppose you define success as:

"The duration category with the highest median of channel-level medians."

The winner becomes 121–180 seconds.

And if you define success as:

"The duration most likely to beat another duration within the same channel."

Our results show no overwhelming winner.

This is a fundamental problem with many online claims about the best YouTube Shorts length.

The answer depends on the question and the measurement method.

Three different research questions

Question 1: Which duration had the highest typical public views across all sampled videos?

Answer: 16–30 seconds.

Question 2: Which duration had the highest channel-balanced median of medians?

Answer: 121–180 seconds in the qualifying channel-level comparison.

Question 3: Does a 20-second video consistently outperform a two-to-three-minute video on the same channel?

Answer: Our sample does not establish that. Among 26 qualifying channels, the comparison split evenly.

These results are not contradictions.

They describe different ways of weighting and comparing a heterogeneous set of videos.

That is why a single "best length" headline is usually less useful than understanding the limits of the data.

Why Different YouTube Shorts Studies Disagree

You may have read that the best Shorts length is 15 seconds.

Or 30 seconds.

Or 50–60 seconds.

There are published studies supporting different ranges.

For example, a 2026 study from quso.ai analyzed 108,138 matured Shorts and reported its strongest median-view category at 11–20 seconds.

An EchoEtch study from May 2026 focused on 1,915 short videos across seven faceless-friendly topic areas and found the strongest results in its 0–15-second category.

A Sera Studio study of 4,452 Shorts appearing on Korean popular charts reported a median duration of 43 seconds, while noting important differences by category.

These studies ask related questions, but they do not use identical samples.

Why the results differ

The datasets vary in:

• Geography.
• Niche.
• Channel size.
• Collection method.
• Minimum views.
• Observation window.
• Whether only popular videos were selected.
• How Shorts were identified.
• Duration-band boundaries.
• View-count definitions.
• Whether every video or every channel received equal weight.

One study may examine recent small-channel uploads.

Another may focus on videos that already became popular.

A third may examine older, mature videos.

Those samples can produce very different rankings.

A crucial measurement change in 2026

YouTube officially changed how public views are counted across formats beginning August 24, 2026.

Under the updated rule, a view is counted when the video begins to play.

YouTube continues to distinguish public views from engaged views and related monetization metrics.

See YouTube's official explanation of engagement metrics.

This matters when comparing older studies with recent data.

Our selected public-view observations were collected from late September through October 2026, after the new rule took effect.

Nevertheless, some underlying videos may have accumulated views under more than one counting regime.

We did not reconstruct view histories or separate pre-change and post-change views for each video.

So results from different periods should not automatically be treated as directly comparable.

The Best Shorts Length Depends on What You're Creating

The research does not support choosing every runtime from one table.

Instead, match the duration to the content.

1–15 seconds: One immediate idea

This range can work when the entire value fits into one quick moment.

Examples:

• A visual transformation.
• A simple before-and-after.
• A surprising comparison.
• A single useful tip.
• A brief reveal.

The danger is removing necessary context just to reach an arbitrary duration.

16–30 seconds: A focused question and answer

This is a useful starting range for short-form videos built around one clear promise.

Examples:

• An unusual historical fact.
• A quick engineering explanation.
• A short psychology insight.
• A financial misconception.
• A brief AI demonstration.

Our pooled study showed the strongest median performance in this category.

But that does not mean every subject should be condensed to 30 seconds.

31–60 seconds: A developed explanation

This range gives more room to establish a problem and deliver a meaningful payoff.

Consider it for:

• Mini tutorials.
• Several connected facts.
• Product comparisons.
• Short case studies.
• Narrated explanations.

The key is ensuring the additional time contributes something useful.

61–120 seconds: A fuller story

Longer short-form videos may be appropriate when viewers need context.

Examples include:

• A historical event.
• A compact documentary.
• An educational demonstration.
• A short mystery with several clues.
• An AI tool walkthrough.

Avoid extending a simple idea merely to reach a longer duration.

121–180 seconds: A complete mini-documentary

This is particularly relevant to documentary, education, and storytelling channels.

Three minutes can provide a beginning, middle, and ending without requiring a conventional long-form format.

Our study found that this category performed competitively in the pooled data and in several channel-level comparisons.

But it was not a guaranteed winner.

The extra time needs to deliver a better viewing experience.

How to Structure a High-Quality YouTube Short

Regardless of duration, the video needs a clear reason for viewers to keep watching.

A practical framework is:

Promise → Development → Payoff

The exact timing should change with the subject.

Example: A 25-second explainer

0–3 seconds: Promise

"How does a city get drinking water when it has almost no natural freshwater?"

3–15 seconds: Development

Briefly explain the relevant technology with concrete visuals.

15–22 seconds: Proof

Show a real mechanism, scale comparison, or meaningful consequence.

22–25 seconds: Payoff

Deliver a clear conclusion that answers the opening question.

Example: A 55-second educational Short

0–5 seconds: Establish the question or problem.

5–20 seconds: Explain the first important detail.

20–40 seconds: Build the explanation using a second example or demonstration.

40–55 seconds: Deliver the conclusion and why it matters.

Example: A 150-second mini-documentary

0–10 seconds: Introduce the central mystery or event.

10–40 seconds: Establish essential context.

40–85 seconds: Explain the first major development.

85–125 seconds: Introduce the most important complication or discovery.

125–150 seconds: Resolve the question and deliver the payoff.

These are original illustrative structures, not pacing formulas proven by the study.

A video with a naturally compelling visual sequence may require a different structure.

The purpose is to make every segment earn its place.

Don't Confuse Public Views With Audience Retention

One limitation of public Shorts research is that views do not reveal how much people actually watched.

A 20-second video and a 150-second video can have the same public view count.

But their watch experiences may be completely different.

That is why YouTube Studio metrics matter.

According to YouTube's official content-performance documentation, relevant Shorts analytics include:

Engaged views: Views where viewers stayed past the initial seconds, excluding loops under YouTube's definition.

Stayed to watch: The percentage of opportunities where viewers watched rather than swiped away.

Average view duration: Average viewing time among the engaged-view audience.

Average percentage viewed: The average percentage watched among that audience.

These metrics help answer questions that public view counts cannot.

Why percentage viewed is not enough on its own

Imagine:

Short A

Duration: 20 seconds.

Average view duration: 18 seconds.

Average percentage viewed: 90%.

Short B

Duration: 120 seconds.

Average view duration: 70 seconds.

Average percentage viewed: 58.3%.

Short A has the higher percentage watched.

Short B has much more viewing time per engaged view.

Neither number alone establishes which video is more successful.

They represent different dimensions of viewing behavior.

Evaluate them alongside the intended content format, audience response, engaged views, and the video's objective.

How to Test the Best Shorts Length for Your Own Channel

A useful test should compare multiple runtimes on the same channel, not just imitate a generic benchmark.

Here's a practical way to do that.

Step 1: Choose three duration ranges

Start with:

• 16–30 seconds.
• 31–60 seconds.
• 61–180 seconds.

If you primarily publish very short clips, add a separate 1–15-second group.

The exact ranges should reflect the kinds of videos you can produce well.

Step 2: Keep the topics reasonably comparable

Don't compare ten-second jokes with three-minute engineering documentaries and then attribute all performance differences to duration.

Try to compare videos that have similar audience intent and subject matter.

For example, an educational channel might test:

• Short explanations of scientific phenomena.
• Slightly longer versions of similar questions.
• More detailed mini-documentaries within the same broad topic area.

Do not publish low-value duplicate content simply to create matched experiments.

Each video should stand on its own.

Step 3: Publish a sufficient number of videos

One unusually successful upload can distort the result.

Aim for several videos per duration group before drawing conclusions.

Our within-channel research required at least three videos in each compared duration category, and even that is a modest sample.

A larger balanced test is generally more informative.

Step 4: Compare results at the same age

Choose a consistent checkpoint.

For example:

Seven days after publication.

Record:

• Public views.
• Engaged views.
• Stayed-to-watch rate.
• Average view duration.
• Average percentage viewed.
• Likes.
• Comments.
• Subscribers gained.

Then repeat the analysis at a later checkpoint.

Do not compare the lifetime views of a two-year-old video against the views of one published yesterday.

Step 5: Use medians instead of relying on your biggest hit

For each duration range, calculate:

• Median seven-day views.
• Median engaged views.
• Median subscriber gains.
• Median stayed-to-watch rate.
• Median viewing duration.

Pay attention to consistency.

A format that produces one enormous viral hit but underperforms on every other upload may not be the best operating model for your channel.

Likewise, a longer format may be valuable even if it earns slightly fewer views, especially when it supports more meaningful storytelling or audience growth.

Step 6: Decide based on your goal

Different objectives favor different evaluations.

There is no universal Shorts length that optimizes all these outcomes simultaneously.

How OverseerOS Can Help You Research Better Short-Form Videos

You don't need to guess every part of your Shorts strategy.

Start by examining real videos and channel patterns.

Step 1: Analyze channels in your niche

Use the OverseerOS Free YouTube Channel Analyzer to inspect public video data.

You can review recent and top-performing videos, recorded duration, views, likes, titles, and publishing patterns.

Look for channels producing a similar style of short-form content.

Pay attention to the topics and formats associated with strong results.

Public analysis cannot reveal competitors' private retention or swipe-away metrics, so treat this as evidence for research rather than a complete performance diagnosis.

Step 2: Study proven short-form openings

Use the OverseerOS AI Hook Generator and Hook Library.

The Hook Library supports short-form filtering so you can study opening patterns from high-view videos rather than relying entirely on generic AI-generated hooks.

Focus on how those videos establish a clear reason to keep watching.

Don't copy another creator's opening word for word.

Use the pattern to create something original for your topic and audience.

Step 3: Write the script around the intended payoff

Before choosing a fixed runtime, define:

• The viewer question.
• The essential context.
• The explanation or demonstration.
• The payoff.
• The visuals required to support the narration.

A topic that can be explained well in 20 seconds doesn't need 90.

A topic that requires two minutes shouldn't necessarily be compressed into 20 seconds merely because one study reported strong views for shorter videos.

Step 4: Build a repeatable content workflow

Save promising ideas and turn them into distinct scripts and planned videos.

Over time, use your own results to identify which durations fit each recurring content format.

The objective isn't to manufacture every Short to the same length.

It's to develop repeatable formats that consistently deliver value to your particular audience.

Why the "Perfect Shorts Length" Is Usually the Wrong Goal

Creators often ask for the perfect runtime because it offers an easy rule.

If every successful Short were 23 seconds, production decisions would become simple.

But our data doesn't support that kind of precision.

The best pooled median appeared in one category.

The channel-balanced ranking favored another.

And the strongest direct comparison between 16–30-second and 121–180-second uploads was evenly split among the 26 qualifying channels.

This suggests a more productive principle:

Different creators and subjects can succeed with different runtimes.

Think of duration as a production choice rather than an independent growth mechanism.

A shorter video may be easier to finish.

A longer video may create a more satisfying explanation.

A tightly edited 45-second video may work better than either.

The data cannot tell you which one will succeed without knowing the content and audience.

That's why your own channel-level tests matter more than treating a platform-wide median as a universal rule.

Important Limitations of This Research

The findings are useful, but they require several qualifications.

1. This is not a random sample of YouTube Shorts

The 16,537 videos came from public data previously encountered and revisited by OverseerOS systems.

Certain channels, subjects, or content types may be overrepresented.

The results are not platform-wide success probabilities.

2. Short-form classification is not verified Shorts-feed eligibility

Our research used the available short-form classification and duration metadata.

We did not independently verify aspect ratios or Shorts-feed distribution for every video.

YouTube's official Shorts classification depends on more than duration alone.

Consequently, this is best understood as a study of short-form YouTube uploads relevant to Shorts-length research, not a validated census of confirmed Shorts-feed videos.

3. We measured public views, not retention

The dataset does not include private average view duration, completion rates, engaged views, or swipe behavior.

A higher median public view count does not automatically indicate a more satisfying video.

4. The videos were not all measured at exactly the same age

We used observations between days 7 and 30, preferring the one closest to day 21.

The median observation age was 18 days.

We performed narrower age-window checks, but some age-related variation remains.

5. Channels contribute different numbers of videos

Some channels publish far more frequently than others.

We addressed this through channel-level and within-channel comparisons, but no weighting method removes every difference between creator populations.

6. Video length is not randomly assigned

Creators choose durations based on their topics, editing styles, and audiences.

That creates confounding.

The observed associations should not be interpreted as the causal effect of adding or removing seconds.

7. We did not control for every video subject

A 20-second entertainment clip and a three-minute educational explanation may attract different audiences.

The research does not isolate duration while holding niche, storytelling quality, language, and audience demand constant.

8. The three-minute category may be compositionally different

Creators choosing 121–180 seconds may produce content unlike creators choosing 15–30 seconds.

This could help explain why longer videos performed competitively in the pooled sample.

We did not prove the cause.

9. Public view counting changed in August 2026

YouTube began counting views from playback start across formats on August 24, 2026.

Our public observations were collected after that change, but some videos may contain accumulated views from different counting periods.

This matters when comparing the findings with older studies.

10. Higher views do not establish higher revenue or audience quality

We did not measure RPM, ad revenue, subscriptions, downstream watch time, or business outcomes.

The best duration for reach may not be the best duration for another objective.

Final Verdict: How Long Should YouTube Shorts Be?

If you need a starting point, test 16–30 seconds. But don't assume longer Shorts are less effective.

In our October 2026 study of 16,537 short-form YouTube videos:

• The 16–30-second category had the highest pooled median views at 36,970.
• The 121–180-second category ranked second at 31,847.
• Extremely short 1–15-second videos did not lead the sample.
• The 16–30-second category beat 31–60 seconds on 57.4% of qualifying paired channels.
• The direct comparison with 121–180 seconds split 13–13 across 26 channels.
• Channel-balanced results did not reproduce a universal short-video advantage.

The research supports testing multiple durations rather than treating one runtime as a requirement.

For a single fact, demonstration, or focused question, 16–30 seconds can be an efficient format.

For a complete explanation, 30–60 seconds may be more appropriate.

For storytelling, education, and mini-documentaries, the full three minutes can be worth exploring.

Your job is to choose the duration that lets the video deliver its promise without unnecessary delay.

Then measure how your audience responds.

Use OverseerOS to research public channel performance, study short-form hooks, and develop original video ideas.

The winning strategy isn't to make every video shorter.

It's to make every second worth watching.

Frequently Asked Questions

What is the best length for YouTube Shorts in 2026?

In an OverseerOS analysis of 16,537 short-form YouTube videos, the 16–30-second category had the highest pooled median views at 36,970. However, within-channel comparisons showed no universal winner, and 121–180-second videos were also competitive. The best duration depends on the content and audience.

Are 15-second YouTube Shorts better than 30-second Shorts?

Not in this pooled sample. Videos lasting 1–15 seconds had a median of 21,913 views, compared with 36,970 for 16–30 seconds. The difference is observational and does not establish that extending a video to 30 seconds will cause higher views.

Are 30-second YouTube Shorts effective?

Yes. The 16–30-second category had the highest pooled median public views among the seven ranges we studied. It is a reasonable starting point for focused videos that can establish a question and deliver a complete payoff quickly.

Are 60-second YouTube Shorts too long?

No. Videos lasting 46–60 seconds had a median of 20,938 public views in our sample. Some subjects require more context than a 20-second clip can provide. The quality and relevance of the additional content matter more than reaching a particular duration.

Do longer YouTube Shorts get fewer views?

Not consistently. In our study, videos lasting 121–180 seconds had the second-highest pooled median views. Within 26 qualifying channels comparing that range with 16–30 seconds, the better-performing category was evenly split.

Can YouTube Shorts be three minutes long in 2026?

Yes. YouTube allows eligible square or vertical Shorts up to three minutes. Standard-channel eligibility for this format began with uploads on or after October 15, 2024. Additional rules apply to copyrighted material, especially Shorts longer than one minute.

Is a two-minute YouTube Short too long?

Not automatically. A two-minute video can be appropriate for a detailed explanation, demonstration, or story. What matters is whether the additional time serves the viewer's expectations. Our research did not establish a universal performance penalty for longer short-form uploads.

What length of YouTube Shorts gets the most views?

The 16–30-second category had the highest pooled median views in our 16,537-video study. But the ranking changed under channel balancing, so this should be treated as a sample result rather than a guaranteed winning duration.

Do shorter YouTube Shorts have better retention?

Shorter videos may be easier to finish, but our research did not measure private retention data. A video's average percentage viewed, stayed-to-watch rate, and average view duration must be examined in YouTube Studio before drawing conclusions about retention.

How long should educational YouTube Shorts be?

Use enough time to establish the question, explain the relevant information, and deliver a satisfying conclusion. A simple insight might fit within 20–30 seconds, while a more detailed demonstration or mini-documentary may require one to three minutes.

How long should faceless YouTube Shorts be?

The appropriate duration depends on the narration, visuals, and complexity of the subject. For a focused fact or quick explanation, 16–30 seconds is a reasonable test. Longer storytelling formats can also work when the additional detail provides meaningful value.

Is 20 seconds the sweet spot for YouTube Shorts?

Twenty seconds falls inside the strongest pooled duration category in our study. However, the analysis does not establish that exactly 20 seconds is optimal or that other durations are inferior.

Does YouTube favor shorter Shorts?

Our observational study cannot establish that YouTube's recommendation system directly favors one video length. Public view differences also reflect channel size, subject, audience demand, and other factors.

Why do different studies recommend different Shorts lengths?

Different studies use different creator populations, video ages, view-count definitions, thresholds, duration ranges, and weighting methods. A study of popular videos may produce a different answer from a study of recent uploads across smaller channels.

How should I test YouTube Shorts length?

Create several original, reasonably comparable videos in multiple duration groups. Compare performance at the same video age using views, engaged views, stayed-to-watch rate, average view duration, and subscribers gained. Use medians rather than judging the entire format from one viral upload.

Should I make every YouTube Short under 30 seconds?

No. A fixed limit can be useful for a specific recurring format, but it is not supported as a universal rule. Use the duration required to deliver the video's promise clearly and efficiently.

Can three-minute Shorts earn money?

Eligible Shorts that appear in the Shorts feed can participate in Shorts revenue sharing when the channel meets the relevant YouTube Partner Program requirements and other monetization policies. Copyright and Content ID restrictions are particularly important for Shorts longer than one minute.

What metric matters most when choosing Shorts length?

No single metric is sufficient. Use public views to understand reach, but also examine engaged views, stayed-to-watch rate, average view duration, average percentage viewed, and whether the video achieves your intended goal.