The best time to post on YouTube: two studies, one overlap, and the thumbnail that matters more

Post long-form video on Sunday morning, around 9 to 10 a.m. your audience's time, and post Shorts on Friday between 4 and 7 p.m. That is where the two largest public datasets, Buffer's 1.8 million videos and SocialPilot's 301 thousand, actually overlap. Everything else they claim, they claim against each other, and this article will show you exactly where, because a best time you can only believe by not reading the second study is not a best time. Then it will show you the thing both studies quietly agree moves more views than any clock: what your video looks like before anyone clicks it.

What the two studies actually found

Buffer measured median engagement across 1.8 million videos and Shorts, hour by hour. Its single best long-form slot is Sunday at 10 a.m., with an engagement factor of 0.95, followed by Sunday at 9 and Friday at noon. Its strongest long-form days are Sunday, Tuesday and Monday, its weakest are Wednesday and Thursday, and its floor is Thursday at 2 p.m. at 0.11. SocialPilot measured engagement per view across 301 thousand videos from 27 thousand channels and came back with weekday afternoons, 2 to 4 p.m., Wednesday through Friday, while also finding that Sunday 8 to 10 a.m. is its highest single long-form window. Read that twice: SocialPilot's best median day is roughly Buffer's worst. On Shorts they converge completely, Friday late afternoon into evening, with Buffer's top three slots at Friday 4, 6 and 7 p.m.

Long-form: mornings, and the Sunday nobody wanted

The morning finding makes intuitive sense once you stop thinking about when people watch and start thinking about when videos accumulate. A long-form video published at 9 a.m. spends the whole day being surfaced, saved and half-watched, and walks into the evening couch session with hours of watch signals already banked. One published at 9 p.m. meets the same couch with nothing. Sunday morning wins on top of that because the competition thins out, most channels publish on weekday schedules set by office habits rather than audience ones.

Shorts: the opposite end of the day

Shorts invert the logic because nobody queues a Short. It is watched in the scroll or not at all, so it wants to publish directly into the scroll's rush hour: after work, end of week. Both datasets put the peak at Friday between 4 and 7 p.m., with Saturday and Thursday behind it, and SocialPilot adds a lunchtime shelf from 12 to 3. If you run both formats, that is two different publishing calendars, and merging them into one is the single most common scheduling mistake a small channel makes. The material for Shorts is a different question from the clock, and the tools that mine what already works in your niche are a better use of an afternoon than any further study of the calendar.

What YouTube itself says

YouTube's help centre describes an audience report in YouTube Studio called "When your viewers are on YouTube", a 28-day heatmap of when your actual viewers are online, and says it exists to help you schedule premieres and streams around them. What YouTube does not say anywhere is that publish time is a ranking signal. The recommendation system runs on what viewers do with a video once it exists: whether they click it, how long they stay, whether they come back. Timing can hand a video a better first evening, and a better first evening can compound, but there is no clock inside the algorithm to please. Which means the studies above are measuring audiences, not rules, and your audience is the one 28-day heatmap no study has seen.

Why every guide you have read gives a different hour

Because they are all downstream of a handful of datasets like the two above, filtered through rewrites until the caveats fall off. A median across 27 thousand channels becomes "post at 2 p.m.", loses the words median and engagement on the way, and arrives at your desk as a rule. Then a second guide does the same to the other dataset and contradicts the first, and both put the current year in the title. The honest version fits in three sentences: mornings for videos, evenings for Shorts, Sunday and Friday are the safe extremes. Beyond that the data disagrees with itself, and anyone selling you a more precise hour than that is selling precision they do not have.

The part that moves more than the clock

Buffer's own study, having ranked all 168 hours of the week, ends by saying the ranking matters less than the topic, the title, the thumbnail and whether the video holds attention. SocialPilot claims shifting your window can move first-day views by 20 to 40 percent, and even taking that at face value, a thumbnail regularly moves click-through by more than that on its own. So here is the uncomfortable arithmetic: the hour decides who is standing in front of the shelf, the thumbnail decides who picks the box up. The rest of this article is thumbnails, because after the schedule question is settled in two lines, this is the craft that actually compounds. The principles are the same ones that make a feed post stop a scroll, tightened for a rectangle two centimetres wide.

One face, one feeling, four words

Let a number do the promising

The series look: one channel, recognisable at a glance

Worlds, not screenshots

What a small channel should actually do

• Open YouTube Studio, Analytics, Audience, "When your viewers are on YouTube". If the heatmap is thin because the channel is young, borrow the studies' overlap as a starting point: Sunday morning for videos, Friday evening for Shorts, in your audience's timezone, not yours.
• Publish one to two hours before the peak, not at it. The video needs to be indexed, titled and visible when the wave arrives, not uploading into it.
• Pick two slots and hold them for eight weeks. A slot is a habit you are teaching an audience, and eight data points is the minimum before first-24-hour views tell you anything.
• **Spend the saved deliberation on the thumbnail.** Draft three, shrink them to 120 pixels wide, which is the size of the smallest file YouTube actually serves, and keep the one still legible. Every frame in this article was judged at that size.
• Schedule the upload, then leave it alone. There is no bonus for babysitting the first hour, and YouTube says waiting between upload and publish does nothing at all.

So: Sunday morning, Friday evening, your own heatmap over anyone's study, and the hours you were about to spend on the perfect slot spent on the rectangle people actually click. The same honesty about what data can and cannot tell you applies across every platform's posting-time folklore, and it applies hardest to the best time to post on TikTok, where four studies this size contradict each other on all seven days, and if the videos are one part of a wider presence, how the rest of it gets found is the companion piece to this one.