The best time to post on YouTube, and why your own numbers rarely name one

Why do published best times not transfer to your channel?

A best time to post is a claim about an audience, and the articles that print one are measuring an audience that is not yours. To say anything about a particular hour, a study has to pool a great many videos from a great many channels, across every niche and every time zone, and report what they have in common. The result is a fact about YouTube as a whole. It says very little about when the people who watch your channel open the app.

Then there is the size of the effect. A difference that shows up across a huge pooled sample can be real and still be far smaller than the difference between one of your videos and the next. On a channel of ordinary size, checking properly usually ends the same way: no day is measurably ahead, and what the video was decided how it did.

None of which means timing is irrelevant. It means the question worth answering is narrower: when is your audience around, and can you keep publishing into that window every week?

Start with the report YouTube already gives you

YouTube Studio has one report aimed squarely at this. Open Analytics, then the Audience tab, and find When your viewers are on YouTube. YouTube describes it as showing when your viewers are online across YouTube over the last 28 days.

Read that description carefully, because it is narrower than it looks. It is when your viewers are on YouTube at all, not when they watch you. YouTube pitches it for timing a Premiere or a live stream, where the audience has to be present at the moment, and its tips page suggests using it to work out what time and day suit your uploads. It is the right starting point because it is about your audience rather than everybody’s. It is not proof that one slot will beat another.

A reasonable way to use it: find the band of hours where your viewers are most active, pick a slot at the start of it or a little before, and then stop moving it.

Why a fixed slot matters more than the perfect one

YouTube’s own upload schedule tips open with the point: a consistent, sustainable release schedule is critical to building and meeting what your audience expects. A slot you keep does that. A slot you chase does not, because a channel that moves its upload time every week has taught nobody to expect anything.

A fixed slot also makes your own numbers readable. If every video goes out at the same time on the same days, the thing that varies between them is the video, and you can compare them. If the time moves too, you can never separate the slot from the video. How often to post on YouTube covers choosing a cadence you can hold in a bad week; the slot is the same decision at a smaller scale.

How to test your own weekdays in a spreadsheet

If you want to know whether a weekday really is better for you, a spreadsheet and six rules will tell you without letting you fool yourself. Do the whole thing twice, once for Shorts and once for long cuts.

1. One row per video. Publish date, weekday, format, and views after the first 48 hours. Write the 48-hour figure down two days after each video goes out: it is the number most tied to the slot rather than to how long a video has been around. If you are starting from old videos, current views are the fallback, and a weaker one, because an older video has simply had longer.
2. One time zone. Record weekdays in the time zone most of your audience lives in, and keep it the same all the way down. A video that goes out late on Sunday in one zone is a Monday video in another.
3. Buckets of at least two. Group the rows by weekday and drop any weekday with fewer than two videos, because one video is not a slot. If fewer than two weekdays are left, stop: there is nothing to compare yet.
4. Medians, not averages. Use MEDIAN() for each weekday. View counts are skewed enough that one video that took off drags an average somewhere none of the others have ever been.
5. The spread. Use QUARTILE(range, 1) and QUARTILE(range, 3) for each weekday. Between those two numbers sit the middle half of that day’s videos, and the gap is the interquartile range.
6. A winner only when the ranges do not overlap. Take the weekday with the highest median and the one with the lowest. Only if the best day’s first quartile is above the worst day’s third quartile is the gap between them bigger than the spread within them. Otherwise the honest answer is that no day is better yet.

Two cautions. The test is conservative on purpose, so it will say “cannot tell” to some real differences that more videos would confirm. That is the right way round to be wrong, because acting on a false best day costs weeks of posting into the wrong slot. And with three or four videos on each day, even a clean result is a reason to try that slot deliberately, not a law.

What Runsheet does with the same question

The same rules, applied in code and published on the methodology page:

• Weekday only, never hour.
• Nothing at all until the channel has at least 8 published videos with a date.
• A weekday with fewer than two videos is dropped, and at least two weekdays must remain.
• Medians, and a best day named only when its interquartile range clears the worst day’s. The test wants at least three videos on each side, so a day with two can appear on the chart but can never be named the winner.
• It is written to use each video’s first 48 hours where it has them and to fall back to current totals where it does not. Today the analytics screen hands it current totals, which is the weaker measure of the two, for the reason in the spreadsheet method above.

Most of the time the answer it gives is that no day is meaningfully better. That is a real finding, and a good reason not to reorganise your week. It sits on the analytics screen beside the other findings, each showing how many videos it rests on.

Why not the best hour?

Because a week has 168 hours. Split a typical catalogue across 168 hourly buckets and the winning hour is decided by one video, which is a best time to post built on a single upload. Weekdays give seven buckets, and even that asks a lot of a channel with a few dozen videos. An hourly answer from a small channel’s own data is not an answer; it is the last video that happened to do well.

Shorts and long cuts, separately

Run the exercise twice, never once over both. A Short is distributed by a feed that decides within hours, while a long cut can still be picking up views months later, so their first 48 hours mean different things. Mix them and your best day turns out to be whichever day you happened to post the most Shorts on. Shorts against long cuts goes into how to compare the two formats honestly.

The time zone trap

When you schedule in Studio, YouTube asks for a date, a time and a time zone. Check the time zone every time rather than assuming it is yours, especially if you schedule while travelling. The same page notes that the date on the watch page is based on Pacific time, so a video set to go out early in a zone ahead of Pacific can show the previous day’s date. That is not a failure, but it will quietly put videos in the wrong row of a weekday spreadsheet if you copy dates off the watch page. Take them from Studio, in one time zone.

If you want to sketch the rhythm before committing to a slot, the cadence planner is free and needs no account.