How Long Should a YouTube Video Be? Data From 1,257 Recent Videos

There is no universal perfect YouTube video length.

But there is a useful range where recent long-form videos in our dataset performed much better than shorter or extremely long uploads.

OverseerOS analyzed 1,257 recent long-form YouTube videos across 245 channels, controlling for channel size and video age.

The strongest result:

Videos between 30 and 60 minutes had the highest median size-and-age-adjusted reach at 1.40x expected performance.

But that does not mean every creator should make 45-minute videos.

The broader pattern was more useful:

• 3 to 8 minutes: 0.23x expected reach
• 8 to 12 minutes: 0.88x
• 12 to 20 minutes: 1.26x
• 20 to 30 minutes: 1.20x
• 30 to 60 minutes: 1.40x
• 60 to 120 minutes: 1.14x
• 120+ minutes: 0.25x

So the strongest zone in this sample was not the commonly repeated idea that every YouTube video should be 8 to 12 minutes.

It was much broader:

Recent videos between roughly 12 and 60 minutes produced the strongest channel-weighted reach after we adjusted for channel size and video age.

There was another important twist.

The length with the highest reach was not the length with the highest visible engagement rate.

Videos between 8 and 12 minutes had the highest median engagement at 3.85%, while 30-to-60-minute videos had the highest adjusted reach.

That means "best video length" changes depending on what you are actually optimizing.

Key Findings

*Channel-weighted share of videos exceeding the expected reach benchmark for videos of similar channel size and age.

Other findings:

• Videos analyzed: 1,257
• Channels represented: 245
• Median video length: 21.7 minutes
• Middle 50% of runtimes: 13.8 to 35.9 minutes
• Median video age at observation: 15.5 days
• In a within-channel comparison of 86 channels, the longer half of each channel's videos outperformed the shorter half in 62.8% of channels
• The median longer-vs-shorter reach ratio was 1.15x
• When we restricted the comparison to channels whose normal runtime was between 8 and 60 minutes, longer videos still won in 59.7% of channels, but the median advantage fell to only 1.06x

That last finding is important.

Longer videos had an advantage often enough to be interesting.

But the advantage was nowhere near large enough to justify padding a 12-minute idea into 30 minutes.

How Long Should a YouTube Video Be?

For most creators, the best answer is:

Make the video as long as the idea can remain valuable and compelling, then judge that runtime against your own recent videos of similar type.

If you want a public-data starting benchmark from this study:

Under 8 minutes

Weakest recent reach in this long-form cohort.

8 to 12 minutes

Much stronger engagement, but still slightly below the expected reach benchmark.

12 to 20 minutes

Strong balance of reach and engagement.

20 to 30 minutes

Also strong, with slightly higher raw relative reach.

30 to 60 minutes

Highest adjusted reach in the primary analysis.

60 to 120 minutes

Still above expected reach, but with a smaller and more variable sample.

Over 2 hours

Much weaker adjusted reach in this cohort.

That does not create a universal formula.

It gives you a better starting hypothesis.

How We Analyzed 1,257 Recent YouTube Videos

The purpose of this study was not to ask:

Which video length has the highest raw views?

That would be a weak comparison.

A 40-minute video published by a 5-million-subscriber channel is not directly comparable with a 12-minute video from a channel with 20,000 subscribers.

Older videos have also had more time to accumulate views.

So we built the analysis around relative recent reach.

The primary cohort

We used public YouTube observations from OverseerOS Channel Analyzer research data.

Every qualifying video had:

• a known publication date
• a known video duration
• positive public views
• a same-day positive public subscriber snapshot
• an observation between 7 and 30 days after publication

We then kept no more than the 20 most recent captured uploads per channel before selecting the long-form videos.

The final cohort contained:

1,257 long-form videos from 245 channels

The observations were captured between:

August 11 and August 20, 2026

The qualifying videos were published between:

July 14 and August 12, 2026

What counted as long-form?

For this study:

long-form = more than 3 minutes

Videos of 3 minutes or less were excluded.

We did that because short-form viewing behavior is different enough that combining it with long-form would make the duration comparison less useful.

The video-length buckets

We grouped qualifying videos into:

• 3 to 8 minutes
• 8 to 12 minutes
• 12 to 20 minutes
• 20 to 30 minutes
• 30 to 60 minutes
• 60 to 120 minutes
• 120+ minutes

Those ranges are descriptive research buckets.

They are not official YouTube categories.

Why We Did Not Rank Video Length by Raw Views

Imagine two videos.

Video A

• 20 minutes
• 40,000 subscribers
• 60,000 views

Video B

• 35 minutes
• 2 million subscribers
• 300,000 views

Raw views say Video B performed five times better.

But relative to channel size:

Video A:

150% of subscriber count

Video B:

15%

That changes the interpretation completely.

Our earlier YouTube views-to-subscriber ratio study also found that normal relative reach changes dramatically with channel size.

So simply comparing views across video lengths would risk measuring:

which channels are bigger

rather than:

which runtimes are associated with unusually strong recent reach.

The Adjusted Reach Index

We created a simple normalization for this study.

First, each video received a public:

views-to-subscriber ratio

Then we identified two major contextual variables:

1. channel size
2. video age

Channel-size bands

• Under 10K subscribers
• 10K to 99K
• 100K to 999K
• 1M+

Video-age bands

• 7 to 13 days
• 14 to 20 days
• 21 to 30 days

We calculated the median reach for every size-and-age combination.

Then each video's performance was divided by the expected reach of its comparison group.

So:

Adjusted Reach Index = observed relative reach ÷ expected reach for same size and age group

A score of:

1.0x

means the video was approximately at the benchmark.

2.0x

means roughly twice the benchmark.

0.5x

means roughly half.

Why age matters

A 28-day-old video has had much more time to accumulate views than a video observed on day eight.

Without age adjustment, longer-video categories could appear stronger simply because they happened to contain older videos.

The adjustment reduces that problem.

Why we still weighted channels

One prolific creator can publish many videos.

We did not want one channel with 18 uploads to carry six times the influence of a channel with three.

So after calculating each video's adjusted reach:

1. we calculated the median within each channel and duration bucket
2. then calculated the median across those channels

The primary results are therefore channel-weighted.

Finding 1: 30-to-60-Minute Videos Had the Highest Adjusted Reach

This was the strongest descriptive result.

Median adjusted reach:

1.40x

for videos between 30 and 60 minutes.

Those videos came from:

• 76 channels
• 225 recent uploads

Their median views-to-subscriber ratio was:

14.6%

And on a channel-weighted basis:

60.1%

of the videos in that duration group exceeded the expected reach benchmark for their size and age.

That was the highest outperformance rate of any duration bucket.

This does not mean 45 minutes is the magic number

There are several possible explanations.

Longer videos may disproportionately appear in:

• documentaries
• commentary
• deep dives
• investigations
• storytelling formats
• podcasts
• established content systems

Those content types may differ from shorter videos in many other ways.

The duration itself may not be what created the reach.

This is observational data.

The useful conclusion is narrower:

Videos in the 30-to-60-minute range were fully compatible with strong reach and actually produced the strongest adjusted median in this recent cohort.

That matters because creators are often afraid that a video is "too long" simply because it crosses 20 or 30 minutes.

The data does not support that fear as a universal rule.

Finding 2: 12 to 30 Minutes Was Also Strong

The result did not depend on one 30-to-60-minute bucket.

12 to 20 minutes

• 111 channels
• 330 videos
• 1.26x adjusted reach
• 52.6% above expected
• 3.52% median engagement

20 to 30 minutes

• 91 channels
• 300 videos
• 1.20x adjusted reach
• 55.7% above expected
• 3.55% engagement

These were the largest two video samples in the study.

Together they represent:

630 recent long-form videos

So the strongest generalizable pattern was not:

Make a 45-minute video.

It was:

The 12-to-60-minute range consistently held up well across several separate duration buckets.

That is a much stronger finding.

Finding 3: 3-to-8-Minute Videos Had the Weakest Reach

The shortest long-form group performed very differently.

Videos between 3 and 8 minutes had:

0.23x median adjusted reach

Only:

30.2%

were above the expected benchmark on a channel-weighted basis.

Median views-to-subscriber ratio:

4.8%

Median visible engagement:

2.25%

All were the weakest or among the weakest results in the primary table.

But this is exactly where causal caution matters.

We cannot say:

Turning a 6-minute video into a 12-minute video will increase its reach.

Those can be completely different types of content.

A short:

• tutorial
• update
• clip
• explainer
• answer video

may serve a different viewer intent than a 25-minute investigation.

The correct finding is:

Recent 3-to-8-minute long-form videos in this sample substantially underperformed the size-and-age-adjusted reach of the longer groups.

That is worth investigating.

It is not permission to add six minutes of filler.

Finding 4: Crossing 8 Minutes Was Not a Magic Performance Rule

The 8-minute mark attracts enormous attention in YouTube strategy discussions.

Our data did show a large descriptive difference:

3 to 8 minutes

Adjusted reach:

0.23x

8 to 12 minutes

Adjusted reach:

0.88x

That is a substantial change.

But it does not establish an 8-minute algorithm threshold.

To make the comparison slightly cleaner, we isolated channels that had qualifying videos on both sides:

• at least one 3-to-8-minute video
• at least one 8-to-12-minute video

Only 11 channels qualified.

Among those channels:

81.8%

had higher adjusted reach in their 8-to-12-minute group.

The median 8-to-12 versus under-8 ratio was:

1.87x

Interesting?

Yes.

Enough to claim that 8 minutes itself causes higher distribution?

Absolutely not.

Eleven paired channels is a small sensitivity check, and the videos themselves can differ in topic, intent, and production.

So do not turn:

8 minutes is commercially meaningful

into:

YouTube's recommendation system rewards exactly 8 minutes.

Those are different claims.

Finding 5: 8-to-12-Minute Videos Had the Highest Engagement

This was one of the most useful contradictions in the study.

If we ranked the duration groups by adjusted reach:

30 to 60 minutes won.

If we ranked them by visible public engagement:

8 to 12 minutes won.

Median engagement rate:

This shows why there cannot be one meaningful answer to:

What is the best YouTube video length?

Best for what?

Best observed reach

30 to 60 minutes.

Best visible engagement

8 to 12 minutes.

Strong balance

12 to 30 minutes.

Best for your own channel

Still unknown until you analyze your retention and viewing behavior.

Our YouTube engagement rate benchmark explains why engagement and reach should not be treated as the same thing.

Finding 6: Longer Videos Beat Shorter Videos on 63% of Matched Channels

Duration buckets still compare different channels.

So we ran a second test.

Instead of asking:

Do channels publishing 30-minute videos outperform channels publishing 15-minute videos?

we asked:

Within the same channel, did its longer recent videos outperform its shorter recent videos?

We isolated channels with at least:

6 qualifying long-form videos

There were:

86 channels.

For every channel:

1. We found its median video duration.
2. We split its videos into a shorter half and a longer half.
3. We compared median adjusted reach.

Typical runtime of the shorter half:

19.4 minutes

Typical runtime of the longer half:

25.6 minutes

Median adjusted reach:

• Shorter half: 0.94x
• Longer half: 0.95x

Those headline medians look almost identical.

But channel-by-channel:

54 of 86 channels

had better reach on their longer half.

That is:

62.8%

The median longer-to-shorter reach ratio was:

1.15x

So the typical ratio favored longer videos by around 15%.

This is much weaker than "longer videos always win"

And that is exactly why this analysis is useful.

If longer duration were overwhelmingly powerful, we would expect something like:

• 90% of channels winning with longer videos
• 2x or 3x median differences

We did not see that.

We saw:

a modest tendency toward better relative reach on the longer videos from the same channels.

Useful signal.

Not a magic formula.

Finding 7: The Within-Channel Advantage Shrunk After Removing Extreme Runtime Channels

Some YouTube channels naturally produce very long content.

Others normally stay near 10 minutes.

So we repeated the paired analysis only for channels whose own median runtime sat between:

8 and 60 minutes

That left:

72 channels

Typical shorter group:

18.2 minutes

Typical longer group:

24.6 minutes

Median adjusted reach:

• shorter: 0.96x
• longer: 0.96x

Longer videos won on:

59.7% of channels

Median longer-to-shorter ratio:

1.06x

That is only about a:

6% median advantage.

This is a critical result.

Once we tighten the comparison, the evidence for "longer is better" becomes much weaker.

The conclusion should therefore be:

Length can matter, but content quality, topic, format, and audience fit are likely more important than adding minutes for their own sake.

Finding 8: Videos Longer Than Two Hours Fell Sharply

The 60-to-120-minute group still performed reasonably well.

Median adjusted reach:

1.14x

Channel-weighted share above expected:

51.2%

But videos at:

120 minutes or longer

fell to:

0.25x

Only:

30.2%

were above expected.

Their median views-to-subscriber ratio was:

3.5%

That is a large descriptive decline.

Again, duration may not be the cause.

The 120+ minute bucket can contain fundamentally different content types from a 20-minute edited video.

So this finding should not be interpreted as:

YouTube suppresses videos after two hours.

The study cannot establish that.

It tells us:

Extremely long recent uploads in this cohort had substantially weaker relative reach than the 12-to-120-minute groups.

Finding 9: The Main Pattern Survived Different Recency Cutoffs

Our primary study keeps up to:

20 recent captured uploads per channel.

Maybe the result is an artifact of that specific decision.

So we reran the adjusted-reach analysis using:

• newest 10
• newest 20
• newest 30

Here are the important results.

The 60-to-120-minute result moved considerably because the group was smaller.

But three patterns remained unusually consistent:

3 to 8 minutes stayed weak

0.20x to 0.31x.

30 to 60 minutes stayed strong

1.37x to 1.49x.

120+ minutes stayed weak

0.25x to 0.32x.

That makes those patterns less likely to be caused solely by choosing exactly 20 recent uploads.

The Median Long-Form Video Was 21.7 Minutes

Another useful benchmark came from the runtime distribution itself.

Across all 1,257 qualifying long-form videos:

25th percentile

13.8 minutes

Median

21.7 minutes

75th percentile

35.9 minutes

So half of the recent long-form videos in this sample lasted roughly:

14 to 36 minutes.

That is important context for creators who still think of 10 minutes as the default long-form runtime.

Within this research corpus, the typical long-form upload was already substantially longer.

Again, this is not a census of all YouTube.

It describes the recent analyzed-channel cohort.

Is 5 Minutes Too Short for a YouTube Video?

Not inherently.

A five-minute video can be exactly right for:

• a narrow tutorial
• a single answer
• a quick update
• a concise explanation

You should not add ten minutes simply because a benchmark says longer videos had stronger reach.

But in our recent dataset, the 3-to-8-minute group had the weakest adjusted reach among normal long-form duration buckets.

So if your videos consistently sit below eight minutes and underperform your channel baseline, duration is one variable worth testing.

The experiment should be:

Can I develop stronger ideas that genuinely justify more depth?

Not:

How do I add five minutes of filler?

Is 8 Minutes a Good YouTube Video Length?

Eight minutes can be perfectly appropriate.

But the data does not support treating it as a universal performance target.

The 8-to-12-minute group had:

• 0.88x adjusted reach
• 47.9% above expected
• 3.85% engagement

That makes it especially interesting for creators optimizing for concise, highly interactive videos.

But longer ranges produced stronger adjusted reach.

Is 10 Minutes a Good YouTube Video Length?

Yes, depending on the idea.

Ten minutes falls inside the 8-to-12-minute bucket, which produced the highest median visible engagement rate in the study.

But it did not produce the highest adjusted reach.

If a topic naturally needs:

17 minutes

cutting it to exactly 10 because someone told you that 10 minutes is optimal makes little sense.

Duration should follow the viewer promise.

Is 15 Minutes a Good YouTube Video Length?

Fifteen minutes sat inside one of the strongest-performing groups.

The 12-to-20-minute bucket produced:

• 1.26x adjusted reach
• 52.6% above expected
• 3.52% engagement

It also had the largest channel representation:

111 channels

So 12 to 20 minutes was one of the most defensible "balanced" ranges in this dataset.

Is 20 Minutes Too Long for YouTube?

No.

The study gives no evidence that crossing 20 minutes automatically hurts performance.

Videos between:

20 and 30 minutes

had:

• 1.20x adjusted reach
• 55.7% above expected
• 3.55% engagement

The median long-form video in the entire sample was already:

21.7 minutes.

Twenty minutes is not unusually long in this research cohort.

Are 30-Minute YouTube Videos Too Long?

Not according to this dataset.

The strongest adjusted-reach category was:

30 to 60 minutes

with:

1.40x expected reach.

Around:

60.1%

of videos in that group exceeded their expected size-and-age benchmark on a channel-weighted basis.

That does not mean longer caused the performance.

It does show that strong public reach is clearly compatible with 30+ minute videos.

Are Hour-Long YouTube Videos Too Long?

Not necessarily.

The 60-to-120-minute group still generated:

1.14x adjusted reach

in the primary analysis.

But the result was less stable in sensitivity tests because fewer channels published videos in this range.

So an hour-long video can work extremely well when the audience and format support it.

It should not become the default simply because longer videos sometimes accumulate more total watch time.

Does YouTube Favor Longer Videos?

This study cannot establish that.

We observed an association between certain longer duration ranges and stronger public reach.

But duration is entangled with:

• topic
• niche
• viewer intent
• format
• creator skill
• production quality
• storytelling
• audience loyalty
• packaging

A 45-minute documentary and a five-minute tutorial are not the same product with a different runtime.

So it would be inaccurate to convert these results into:

YouTube's algorithm favors long videos.

The data supports a more careful statement:

Longer videos were not disadvantaged in this recent sample, and the 12-to-60-minute groups produced stronger adjusted reach than the shortest long-form group.

That is as far as the evidence takes us.

Reach and Retention Are Not the Same Metric

This study uses public data.

It cannot see another creator's:

• average view duration
• average percentage viewed
• audience retention graph
• watch time
• impressions
• click-through rate
• unique viewers
• returning viewers

Those metrics are critical to deciding the correct runtime for your own channel.

Imagine:

Video A

10 minutes long

Average percentage viewed:

70%

Average watch time:

7 minutes

Video B

30 minutes long

Average percentage viewed:

35%

Average watch time:

10.5 minutes

Which is better?

Video A retains a much higher percentage.

Video B generates more minutes watched per view.

Now add:

• viewer satisfaction
• impressions
• click-through rate
• total views
• revenue

and the answer can change again.

That is why runtime cannot be optimized from one percentage.

The Best YouTube Video Length Depends on the Idea

A strong runtime is usually an output of the video concept.

Consider these examples.

Narrow tutorial

How to Remove Background Noise in Premiere Pro

Maybe the complete answer takes six minutes.

Making it 25 minutes could destroy the viewer experience.

Investigation

Why This Billion-Dollar Company Suddenly Collapsed

The viewer may happily watch 35 minutes if the story continually develops.

Compressing it into eight minutes could remove:

• context
• evidence
• narrative
• stakes
• explanation

Challenge

I Tried Building a Business in 30 Days

The premise contains:

• progression
• setbacks
• results
• time
• narrative

It may naturally justify a longer runtime.

The mistake is selecting:

20 minutes

before asking:

How much story actually exists?

A Better Framework Than "Make Videos 10 Minutes Long"

Use four questions.

1. How much promise is in the idea?

A narrow promise usually needs less runtime.

A broad promise can support more depth.

2. Does every section advance the payoff?

Every additional minute should do at least one of these:

• answer a question
• introduce evidence
• create stakes
• progress the story
• reveal information
• resolve tension
• create a new justified question

If it does none of them, duration has probably become padding.

3. What does your retention say?

Your own recent analytics should eventually override public benchmarks.

Compare videos of similar:

• topic
• format
• audience
• length

Look for the runtime your audience actually sustains.

4. What happens to reach?

Do not optimize only for retention percentage.

Watch:

• total views
• relative views
• views-to-subscriber ratio
• velocity
• watch time

A longer video can have lower percentage viewed and still be a stronger overall performer.

The 12-to-30-Minute Range May Be the Best Starting Point for Testing

If a creator had no historical data and asked:

Where should I begin testing long-form?

the most defensible broad range from this study would be:

12 to 30 minutes

Why not simply recommend 30 to 60?

Because the 12-to-30 range had:

• larger samples
• strong adjusted reach
• strong engagement
• broad channel representation

12 to 20 minutes

330 videos
111 channels
1.26x adjusted reach
3.52% engagement

20 to 30 minutes

300 videos
91 channels
1.20x adjusted reach
3.55% engagement

That is a strong combination.

Then let your actual audience determine whether you should move:

• shorter
• longer
• deeper
• tighter

The benchmark starts the experiment.

It should not end it.

How to Find the Best Video Length for Your Own Channel

Here is the experiment I would run.

Step 1: Take your last 20 long-form videos

Record:

• runtime
• views
• views relative to channel baseline
• average view duration
• average percentage viewed
• 30-second retention
• total watch time

Step 2: Create three duration groups

Use your own distribution.

For example:

• shorter than 15 minutes
• 15 to 25 minutes
• longer than 25 minutes

Do not copy our buckets if they do not fit your channel.

Step 3: Compare medians

Do not let one viral upload decide the answer.

Calculate the median for each group.

Step 4: Look for the tradeoff

Ask:

Does reach increase as runtime increases?

Does retention percentage collapse?

Does absolute watch time increase?

Does engagement change?

Does production quality drop because longer videos take too much effort?

Step 5: Compare within the same content type

Do not compare:

six-minute tutorials

against:

50-minute documentaries

and conclude the difference came from runtime.

Compare like with like whenever possible.

Step 6: Test deliberately

If 15-to-20-minute videos are working:

try stronger ideas that naturally justify:

20 to 25 minutes

Then measure.

Do not add minutes artificially.

How to Research Competitor Video Length With OverseerOS

You can apply the same logic to competitor research.

1. Find channels actually breaking out

Use the OverseerOS Viral Channel Finder to identify emerging channels producing unusual recent public performance.

Do not begin with the biggest creator in the niche.

Start with the strongest current evidence.

2. Analyze their recent uploads

Open the strongest channels with the AI YouTube Channel Analyzer.

Compare:

• recent videos
• durations
• views
• breakout videos
• recurring topics
• title patterns

You are looking for relationships.

Not one "ideal" number.

3. Compare outliers with normal videos

If most of a competitor's videos are:

12 to 18 minutes

but its strongest recent outliers repeatedly sit around:

25 to 35 minutes

that is strategically interesting.

The reverse can also happen.

4. Ask what changed besides length

Did the longer outlier also have:

• a stronger topic?
• a bigger story?
• a different thumbnail?
• more evidence?
• a broader audience promise?

Duration is rarely the only variable.

5. Build the pattern, not the runtime

Use a Channel Blueprint to identify repeatable patterns behind the channel's strongest videos.

The goal is not:

Their viral videos are 31 minutes, so mine should be 31 minutes.

The goal is:

Their biggest ideas justify deeper videos, and those deeper videos repeatedly outperform.

That is an actual strategy.

Video Length vs Views-to-Subscriber Ratio

Views-to-subscriber ratio asks:

How large was the video's public view count relative to the channel's subscriber count?

This is useful because raw views can hide smaller-channel outliers.

But channel size strongly changes the normal ratio.

That is why this study adjusts the benchmark by subscriber tier before comparing duration.

If you want the full channel-size breakdown, see our YouTube views-to-subscriber ratio research.

Video Length vs Engagement Rate

Engagement asks:

How frequently did viewers visibly like or comment?

The best reach bucket and best engagement bucket were different.

That is useful evidence that creators should not optimize one metric blindly.

Our full YouTube engagement benchmark study breaks down likes and comments separately.

Video Length vs VPH

Views per hour asks:

How quickly is the video accumulating views?

A long video can have:

• high engagement
• low velocity

or:

• moderate engagement
• extremely high velocity

These metrics describe different parts of performance.

Our YouTube VPH study found that raw velocity itself also becomes misleading unless compared with the creator's normal baseline.

A useful performance system therefore looks at:

No one metric gives you the full answer.

YouTube Video Length Benchmark Cheat Sheet

Based on this study:

If you want one sentence:

The data suggests that 12 to 60 minutes can support very strong long-form reach, with 12 to 30 minutes offering the strongest broadly sampled balance, but your actual ideal length should be determined by what the idea can sustain and what your retention data confirms.

The YouTube Video Length Checklist

Before deciding a video is too short or too long, ask:

• Does the topic genuinely justify this runtime?
• Am I adding minutes because the story needs them or because I am targeting a number?
• Does every section advance the promise?
• How do similar-length videos perform on my own channel?
• What happens to average view duration as runtime increases?
• What happens to average percentage viewed?
• Does relative reach improve or decline?
• Am I comparing the same content format?
• Could I cut 20% without losing anything important?
• Could the idea support more depth without adding filler?

The goal is not:

longest possible video.

It is:

longest justified video.

What This Study Does Not Prove

This research has several important limitations.

It is not a random sample of all YouTube

The channels entered the research corpus through public OverseerOS channel-analysis activity.

They may differ systematically from the wider YouTube population.

Duration did not randomly change

Creators chose the runtime.

That means runtime is associated with:

• topic
• format
• niche
• creative style
• creator
• production resources

We cannot isolate duration as the causal variable.

We controlled for size and age, not everything

The Adjusted Reach Index controls for:

• public subscriber-size band
• video-age band

It does not control perfectly for:

• niche
• country
• language
• topic demand
• creator skill
• title quality
• thumbnail quality
• upload frequency
• viewer intent

That is why the within-channel test was important.

Even that cannot remove every confounder.

Public data cannot see retention

This study does not have competitor-level access to:

• audience retention
• watch time
• average percentage viewed
• average view duration
• impressions
• CTR

Those private metrics are essential when optimizing your own video length.

Views are not unique viewers

The public view count can include repeated viewing.

Subscriber count is a same-day snapshot

It is not necessarily the subscriber count the channel had when each video was published.

The longest videos are heterogeneous

A two-hour upload may serve a fundamentally different purpose from a 15-minute edited video.

The 120+ minute result should therefore not be generalized to every extremely long format.

The study measures recent reach, not lifetime performance

Every primary video was observed between:

7 and 30 days after publication.

Some videos may continue accumulating views for months or years.

This study answers:

How did these duration groups perform during a relatively early recent-view window?

It does not answer:

Which length produces the highest lifetime views?

Correlation is not causation

The study cannot prove:

• longer videos cause more recommendations
• 30-to-60-minute videos receive an algorithm boost
• short videos are penalized
• crossing eight minutes increases distribution
• adding runtime improves performance

Those claims require evidence we do not have.

Final Verdict

How long should a YouTube video be?

There is no universal perfect length.

But our analysis of 1,257 recent long-form videos across 245 channels found a clear descriptive pattern.

Median adjusted reach

• 3 to 8 min: 0.23x
• 8 to 12 min: 0.88x
• 12 to 20 min: 1.26x
• 20 to 30 min: 1.20x
• 30 to 60 min: 1.40x
• 60 to 120 min: 1.14x
• 120+ min: 0.25x

The strongest adjusted reach appeared at:

30 to 60 minutes.

The highest visible engagement appeared at:

8 to 12 minutes.

The most defensible broad starting range, based on sample size plus balanced performance, was:

12 to 30 minutes.

And when we compared longer and shorter videos from the same 86 channels, longer videos won:

62.8% of the time

with a median reach advantage of:

1.15x

But that advantage became only 1.06x in a tighter 72-channel sensitivity check.

So the data does not say:

Make every video longer.

It says:

Do not be afraid of length when the idea genuinely earns it.

Your runtime should not be selected because somebody told you YouTube wants:

• 8 minutes
• 10 minutes
• 20 minutes
• 30 minutes

Build the strongest version of the idea.

Cut anything that does not advance the promise.

Then use your actual retention and reach data to learn how much depth your audience rewards.

That is how you find your real ideal YouTube video length.

FAQ

How long should a YouTube video be?

There is no universal ideal runtime. In this OverseerOS analysis of 1,257 recent long-form videos, the strongest size-and-age-adjusted reach appeared in the 30-to-60-minute group, while 12 to 30 minutes offered a strong balance of reach, engagement, and sample size.

What is the best YouTube video length?

The best length is the runtime your idea and audience can sustain without filler. In this study, 30-to-60-minute videos had the highest adjusted reach at 1.40x expected performance, but that does not prove the duration caused the result.

What is the average YouTube video length?

This study is not a census of all YouTube, so it cannot provide a platform-wide average. Among the 1,257 recent long-form videos analyzed, the median runtime was 21.7 minutes and the middle 50% ranged from 13.8 to 35.9 minutes.

Is 5 minutes too short for YouTube?

No. Some ideas only require five minutes. However, the 3-to-8-minute group had the weakest adjusted reach in this particular recent long-form cohort at 0.23x expected reach.

Is 8 minutes a good YouTube video length?

It can be. Videos between 8 and 12 minutes had the highest median visible engagement in the study at 3.85%, although their adjusted reach of 0.88x was below the longer 12-to-60-minute groups.

Is 10 minutes long enough for a YouTube video?

Yes, if the topic can be fully delivered in 10 minutes. Ten minutes falls inside the study's highest-engagement bucket. There is no evidence that every idea should be stretched beyond that point.

Is 15 minutes a good YouTube video length?

Fifteen minutes falls in the 12-to-20-minute group, which produced 1.26x median adjusted reach across 111 channels and 330 videos. It was one of the strongest and most broadly represented duration groups.

Is a 20-minute YouTube video too long?

No. The median long-form video in this study was 21.7 minutes. Videos between 20 and 30 minutes produced 1.20x expected reach and a 3.55% median visible engagement rate.

Is a 30-minute YouTube video too long?

Not according to this dataset. Videos between 30 and 60 minutes produced the highest median adjusted reach at 1.40x expected performance.

Are hour-long YouTube videos too long?

Not automatically. The 60-to-120-minute group still produced 1.14x median adjusted reach, although the sample was smaller and the result was less stable than the 12-to-60-minute groups.

Does YouTube favor longer videos?

This study cannot establish that. Longer videos were associated with stronger recent reach in several duration bands, but length is also related to topic, format, niche, production style, and audience behavior.

Do longer YouTube videos get more views?

Not universally. In the same-channel analysis, the longer half of recent videos outperformed the shorter half in 62.8% of 86 channels, with a median 1.15x reach ratio. The advantage became smaller in a tighter sensitivity test.

Does video length affect YouTube engagement?

It can be associated with different engagement patterns. In this study, 8-to-12-minute videos had the highest median visible engagement at 3.85%, while the highest adjusted reach occurred at 30 to 60 minutes.

Should I make videos longer to increase watch time?

Do not add filler merely to increase runtime. A longer video only creates more potential watch time if viewers continue finding the content valuable enough to watch.

How do I find the ideal video length for my channel?

Group your recent comparable videos by runtime and compare median reach, average view duration, percentage viewed, 30-second retention, and total watch time. Your own audience data should ultimately outweigh platform-wide benchmarks.

What length should a new YouTuber start with?

If you have no channel-specific evidence yet, this study suggests 12 to 30 minutes is a reasonable long-form testing range because it combined strong adjusted reach, solid engagement, and relatively large samples. The idea itself should still determine the final runtime.