Best time to post on YouTube

Does posting time actually matter on YouTube?

Less than almost anything else you could change. I have watched channels move their upload slot by three hours and see nothing, and I have watched a channel double its views on the same slot by fixing the first two seconds of the video. Posting time is the smallest lever on this list, and it is the one people spend the most time arguing about.

There is a decent reason the studies are soft on this. They measure hundreds of thousands of channels with different audiences in different countries, then average them, and an average audience does not exist. Your audience does. After about 30 published videos, YouTube can draw your week for you, and that drawing beats every study on this page.

So use the grid to pick a slot this month, publish on it consistently enough to get clean data, then let your own numbers overrule it. And be honest about which problem you have first, because posting time will not fix a video people scroll past. If the rate on your last ten uploads is flat, the hour is not what is wrong.

Two things worth reading next if you publish beyond YouTube. We keep the same question for TikTok and Reels in a separate tool, because the windows do not carry across platforms. And on TikTok specifically, getting found on TikTok search is a bigger lever than the upload slot, by a distance nobody argues about.

Best time to post YouTube Shorts (and why it differs from long form)

Shorts and long-form do not want the same hour. Buffer separated the two formats for the first time in its July 2026 edition and its Shorts windows sit in the evening on six days out of seven, while its long-form windows cluster in the morning. If you publish both on one schedule, one of them is going out at the wrong end of the day.

The clearest agreement for Shorts is Thursday 7pm to 10pm on your audience clock, where two independent studies name the same three hours. Buffer ranks 7pm, 8pm and 9pm as its top three Thursday slots. SocialPilot names 7pm to 11pm for Thursday. Friday late afternoon is second, and it comes with a caveat you can read in the disagreement section below.

Before you move anything, one Shorts specific note that is worth more than the hour you publish at: the frame matters more than the clock, especially on Shorts. A title and a subscribe button sit on top of your video in the Shorts player, and a caption hidden behind either of them costs you more views than a two hour scheduling change ever will.

Best time to post long-form YouTube videos

Long-form is where the studies fall out with each other. Buffer puts four of its seven best daily slots between 7am and 11am and calls Sunday 10am the best slot of the week. SocialPilot puts every weekday between 1pm and 5pm and lists before 9am among the worst times to publish. Metricool says 9pm gets the most views. Those are three different instructions from three companies looking at real data.

Where they overlap is thin: Sunday around 9am, and Wednesday mid to late afternoon. If you want a defensible default with no channel data of your own yet, take one of those two and hold it for a month.

Where the studies disagree

Most posting time articles average these sources together and print one confident number. That hides the interesting part. Below is every contradiction I found while reading the three source pages, including one source contradicting itself.

Long-form videos

• Mornings or early afternoons. Buffer puts its best long-form slot between 7am and 11am on four days of the seven. SocialPilot names 1pm to 5pm on weekdays and lists before 9am as one of the worst times to publish. Those are close to opposite instructions.
• Late evening. Metricool says 9pm gets the most views for long-form, and Buffer has Monday 10pm as its second best Monday slot. SocialPilot lists after 10pm among the worst times.
• Best day of the week. Buffer ranks Sunday first, then Tuesday, then Monday, and calls Wednesday and Thursday the weakest days. SocialPilot ranks Wednesday, Thursday and Friday as the best. They contradict each other on the same two days.

Shorts

• Sunday evening. The two best Sunday slots for Shorts in the Buffer data are 7pm and 8pm. SocialPilot lists Sunday evenings as a worst window for Shorts.
• Friday, inside one source. The SocialPilot Friday row for Shorts says 7am to 9am or 11am to 1pm, while its own summary calls Friday 4pm to 7pm the strongest single Shorts window. The Buffer Friday slots are 4pm, 6pm and 7pm, which back the summary rather than the row.

None of that makes the studies useless. It makes them a shortlist of candidate slots to test, which is all a published average was ever able to be.

Find your own best time in YouTube Studio (step by step)

YouTube gives you the answer for your channel, free, and it takes about a minute to read.

1. Open YouTube Studio and click Analytics in the left menu.
2. Open the Audience tab.
3. Scroll to the card called When your viewers are on YouTube. It draws your week as a grid, darker where more of your viewers are watching.
4. Check what clock the chart is labelled in before you plan around it, then find your two or three darkest blocks.
5. Publish an hour or two ahead of a dark block rather than into it, so the video has been indexed and tested by the time your people show up. That is the standard habit rather than a published law, and I have not seen anyone show the test behind it.
6. Hold that slot for six to eight uploads before you judge it. One video tells you nothing, and a slot you keep moving can never be measured.

If the card is empty, your channel does not have enough traffic yet. Use the grid above, publish, and come back to it once you have a run of videos behind you.

One more habit worth having: keep a note of upload time next to views, retention and subscribers gained for every video. Ten rows of that spreadsheet is more useful than any article about posting times, including this one.

And keep the order of operations straight while you are in there. Write to the runtime before you worry about the upload slot, because a video that runs 40 seconds longer than the point it is making loses people at every hour of the day.

How this planner works

• Every window comes from a named page. No averaging across sources, no rounding of one study towards another, nothing from a search snippet. Where a source names a single hour, it gets one hour on the grid.
• The grid is on your audience clock, and it does not move when you change timezone. All three sources publish in the audience's local time, and Buffer says so outright, so converting their windows into a different zone would misquote them. What the timezone pickers change is the second clock: the "on your clock" column, which tells you when to press upload where you are.
• Shading counts studies, not confidence. One study naming an hour gets the faintest shade, two gets darker, three darkest. A ring means a source ranked that hour first for the day. Slots where a source contradicts itself are marked, not quietly dropped.
• Claims with no day attached are printed, not plotted. Metricool says 5pm is popular for Shorts without naming a day. Spreading that across all seven days would invent data, so it sits under the grid as text instead.
• Daylight saving is handled by your browser using the standard timezone database. The conversion is measured at today's offsets, so a window that straddles a clock change can be an hour out for a few days a year.
• Sprout Social is missing on purpose. Its March 2026 report covers Facebook, Instagram, LinkedIn, Pinterest, TikTok and X, and not YouTube, so its numbers cannot honestly carry a YouTube label. Hopper HQ is out too, because its page says it synthesises the Buffer study, and counting it would count Buffer twice.

Sources

Three sources, all read on the page rather than in a search result. Publication dates, sample sizes and the exact method are below, because a posting time is only worth as much as the study behind it.

The page names its sample for TikTok and LinkedIn but not for YouTube, and states no timezone, so only its specific day and hour claims are used here. Its day and hour claims are on the grid, its undated general advice is not.

Last verified: July 2026. These pages get rebuilt every year, so the windows are re-checked quarterly and the date above moves with them.