How Long Should a YouTube Video Be in 2026?

Your YouTube video should be as long as you can keep earning the next minute. There is no magic runtime in 2026. Eight minutes is not an algorithm switch, 10 minutes is not automatically ideal, and longer does not mean more valuable.

A useful working rule for long-form videos is this: choose the longest version that can hold retention around 40% or better, then judge it against videos of similar length and format. That is a floor for decision-making, not an official YouTube threshold.

Why there is no perfect YouTube video length

Video length is an output of the idea. A narrow tutorial might need 5 minutes. A documented investigation might need 25. Forcing both into the same runtime makes one incomplete and the other padded.

The viewer does not reward minutes. The viewer rewards progress toward the promise in the title and thumbnail. Every section must answer a question, raise a new one, show evidence, create tension, or deliver a payoff.

YouTube also evaluates more than raw watch time. Its recommendation systems operate in a watch time plus satisfaction era. A video that earns a long view by withholding the answer can still leave the viewer dissatisfied. The better length is the one that delivers fully without making the audience work through filler.

Retention changes as videos get longer

Percentage retention normally falls as runtime increases. That does not make long videos worse. It changes how you judge them.

The 2026 YouTube retention benchmarks use these strong ranges:

The watch-time examples are simple conversions, not separate benchmarks. They show why a longer video can win with a lower percentage: a 20-minute video at 40% produces 8 minutes of average view duration, while an 8-minute video at 55% produces 4 minutes 24 seconds.

But the long video only wins if the extra minutes are worth watching. If extending the 8-minute idea to 20 minutes drops retention to 20%, both versions earn 4 minutes. The viewer just had to navigate more dead weight in the longer one.

How long each YouTube format should be

Use these as starting ranges for planning, not performance statistics. Your own channel data should move the boundaries.

Tutorials: long enough to complete the task. A focused answer may fit in 4-8 minutes. A start-to-finish workflow may need 10-20. Put the immediate answer first, then explain exceptions and detail. Search viewers often leave once their problem is solved, so completion rate is not the only sign of value.

Vlogs: long enough to complete one story. A vlog needs a change between the opening and ending. That could fit in 8 minutes or 25. Cut repeated travel shots, routine transitions, and events that did not change the story.

Commentary and video essays: long enough to prove the argument. Use one claim per section. When a new section only restates the previous point, the script is done. Existing benchmarks place strong 8-12 minute commentary retention at 50-60%, but your topic may justify a much longer essay.

YouTube Shorts: long enough to deliver one payoff. Plan 15 seconds for a single reveal, 30 seconds for a compact explanation, or 60 seconds for a short story with escalation. These are structural templates, not universal targets. Shorts are judged with different viewing behavior, including swipe decisions and loops.

Podcast clips: long enough to make one moment self-contained. Start after the original conversation's setup, add only the context a new viewer needs, and end at the peak. A clip should feel like a complete piece, not a random excerpt from a longer recording.

Before recording, use the words-to-minutes calculator to turn script length into an estimated runtime. It is easier to cut 250 weak words in a document than to rescue them in the edit.

The 8-12 minute ad-revenue myth

The myth starts with a real platform rule: videos at least 8 minutes long can be eligible for mid-roll ads under YouTube's monetization settings. That makes 8 minutes relevant to ad inventory. It does not make 8 minutes a recommendation threshold.

Creators often reverse the logic. They start with a 6-minute idea, add 2 minutes of setup or repetition, and assume the extra ad opportunity makes the video better. The trade is poor when the added material causes viewers to leave before the strongest section.

Revenue per published video matters. So do return viewers, satisfaction, and the trust created when a video respects its title. A mid-roll slot has no value to a viewer who already clicked away.

If the script naturally lands at 7:20, publish the strongest 7:20 version. If it needs 11 minutes, use 11. Runtime should follow the content decision, not lead it.

When longer YouTube videos win

Longer videos have one clear advantage: more potential watch time per view. They work when the additional runtime adds value and the format supports sustained attention.

The topic has real depth. A complete comparison, investigation, documentary, or course needs evidence and context. Compressing it can make the result less useful.

The structure keeps changing. New questions, locations, attempts, examples, or counterarguments reset attention. Long videos fail when one beat continues without development.

The audience chose depth. Subscribers who expect long analysis behave differently from search viewers who need one answer. Compare retention by traffic source and returning-viewer behavior before shortening a format that already works.

The added section increases the payoff. More runtime is justified when it makes the conclusion clearer, more credible, or more emotionally satisfying.

The key is accumulation. A longer video should accumulate evidence, tension, understanding, or story. If it only accumulates words, cut it.

When longer videos kill retention

The title promises one answer. A viewer who clicks "Best Budget Camera for Travel" does not need a history of camera sensors before the verdict.

The introduction explains the video instead of starting it. Greetings, credentials, channel updates, and a roadmap delay the first useful moment. Confirm the click and begin.

Sections exist to hit a word count. Repeated examples and summaries are visible on a retention curve. They create gradual decline or a sharp exit at the transition.

The payoff arrives, then the video continues. Once the natural ending passes, a recap or extended CTA asks satisfied viewers to stay without giving them a reason.

The middle has no new question. Initial curiosity fades. The viewer needs a new gap to close, a complication to resolve, or a concrete next step.

Read the YouTube retention guide for the specific structural patterns behind these drops.

A script-first decision framework

Do not choose 8, 10, or 15 minutes before you know what the video contains. Use this sequence:

1. •Write the title's promise in one sentence.
2. •List the minimum points needed to deliver that promise completely.
3. •Put the strongest proof, result, or tension in the opening.
4. •Draft the script without targeting a runtime.
5. •Cut every section that does not change what the viewer knows, feels, or expects.
6. •Estimate the runtime from the final word count.
7. •Compare that runtime and format with your own retention history.

If comparable videos lose viewers at the same point, edit for that point before recording. A repeated 35% drop in the curve is more useful than generic advice to "make it shorter." Maybe the runtime is fine and your third section is not.

After publishing, compare average view duration, percentage viewed, and curve shape. The retention graph guide explains how to separate a weak opening from a slow middle or false ending.

The answer in one sentence

Make the shortest video that delivers the full promise, unless additional minutes keep creating enough value to earn their watch time.

That rule works for a 20-second Short and a 40-minute documentary. Script first. Cut to the strongest length. Let audience retention show whether the next version should be tighter or deeper.

Test your script's pacing before you record