How-To Guide
What is the best time to post YouTube Shorts?
Buffer says Friday 4 p.m., Adobe says Tuesday, and YouTube says publish time is not known to affect long-term performance. How to test your own window.
By the Superdirector editorial team. YouTube Help pages and the studies cited read November 7, 2026; worked example and figures added November 11, 2026.
At a glance
What is the best time to post YouTube Shorts?
No one hour wins for every channel. Buffer's July 2026 ranking, drawn from 1.8 million Shorts and long-form videos, lists Friday at 4, 6 and 7 p.m. for Shorts but does not clarify whose local time it means, Adobe's May 2025 sample of 24,274 Shorts favors Tuesday, and YouTube says an effect on long-term performance is not known, though uploading when your audience is active can help early viewership. Use your viewer-activity peak and a distinct study slot as two windows. Without a usable report, Friday and Saturday at 4 p.m. in your intended audience's time zone are an illustrative pair, not a verified interpretation of either study's clock. Post one comparable Short in each window every week for five weeks, then wait until the final Short is 28 days old. Keep a window only if its median engaged views are still at least 1.5 times ahead; this small comparison cannot prove timing caused the difference.
- The big studies disagree: Buffer's best day is Friday, Adobe's is Tuesday.
- YouTube says publish time is not known to affect long-term performance; it can help early viewership.
- Test two windows for five weeks and read engaged views at 24 hours and 28 days.
The studies disagree on the same days, and Tuesday and Friday swap places. Original diagram of Buffer's July 2026 ranking and Adobe's May 2025 data, read November 7, 2026.
Make it work for your brand
Find it. Remake it. Post it. Superdirector turns the viral formats in your niche into your brand's carousels, X posts and video scripts, and keeps the calendar full.
The short answer: no universal hour, and YouTube says it matters less than you think
No one hour wins for every channel, and two cited studies point at different days. In Buffer's July 2026 ranking, drawn from 1.8 million YouTube videos with Shorts and long-form pooled, Friday leads for Shorts and 4, 6 and 7 p.m. are the three best hours; Buffer says its hours apply "in your local time" without saying whether that means yours or your viewers', so treat either time-zone interpretation as uncertain. Adobe's numbers, taken in May 2025 from 24,274 Shorts, put Tuesday first for overall views and likes and Saturday at 4 p.m. as the best single day-and-hour slot. YouTube's own FAQ says an effect on long-term performance is not known, although uploading when your audience is most active can help the first hours. Start from your own viewer-activity report, test two windows and keep the one that wins.
The table sets out each source with the method behind its number, because the method is what explains the disagreement.
| Source | Sample and method | What it found | Limit |
|---|---|---|---|
| Buffer, July 24, 2026 | Median engagement of 1.8 million YouTube videos by posting time; Shorts and long-form reported separately | Best day Friday, then Saturday and Thursday; best slots Friday 4, 6 and 7 p.m.; worst days Tuesday and Monday | The sample mixes Shorts and long-form videos, and a Shorts-only sample size is not stated |
| Adobe Express, data of May 22, 2025 | 280 creators and 24,274 of their Shorts, spread over 12 content categories; views, likes and comments | Tuesday most views and likes; Saturday 4 p.m. best slot; Friday and Monday weakest; Shorts posted outside the 8 to 10 a.m. rush beat those posted inside it | A snapshot of 280 creators; Tuesday leads the overall averages, while its niche-by-niche breakdown names Wednesday the standout day in many niches |
| OpusClip data, via Bytecap, April 2026 | When OpusClip users posted clips | Posting peaked on Wednesday and 5 to 8 p.m. UTC | Bytecap says it shows when users posted, not which time caused the most views |
| CoSchedule, updated September 2026 | Cites aggregated data across industries; no sample size stated | Friday afternoons and evenings, 2 to 6 PM | No method to check |
| FlowShorts, March 2026 | A mix of studies | Weekdays 3 to 6 PM EST | Its Wednesday 4 PM matches Buffer's earlier all-video analysis; Buffer's July 2026 Shorts ranking puts Friday at 4 p.m. first |
Day by day, the gap is easier to see. Buffer's Shorts slots cluster in the evening, and Adobe's day verdicts flip on Tuesday and Friday, though both studies rate Monday weak. Buffer lists the three best hours for every day, weak days included: they are the hours to use if a weak day is the only day you can post, not a reason to pick that day.
| Day | Buffer: top three hours for Shorts | Adobe: what its sample shows |
|---|---|---|
| Monday | 8 p.m., 5 p.m., 6 p.m.; among the worst days | Among the weakest days for likes and views |
| Tuesday | 8 p.m., 9 p.m., 7 p.m.; among the worst days | Best day for views and likes |
| Wednesday | 7 p.m., 8 p.m., 9 p.m. | Second to Tuesday overall; the standout day in many niches in its niche breakdown |
| Thursday | 7 p.m., 8 p.m., 9 p.m.; third-best day | The most popular day to post, with 20% of the Shorts |
| Friday | 4 p.m., 6 p.m., 7 p.m.; best day | Among the weakest days |
| Saturday | 7 p.m., 11 a.m., 6 p.m.; second-best day, the only one with a morning slot | 4 p.m. was the best slot for views and likes |
| Sunday | 7 p.m., 8 p.m., 5 p.m. | Not singled out |
Adobe's most common upload time was 8 a.m., yet it found Shorts posted outside the 8 to 10 a.m. rush did better, so popular and best are different questions.
The two studies contradict each other on days. Tuesday tops Adobe's overall averages and lands near the bottom of Buffer's, and Friday does the opposite. Postminion points to four differences that make them hard to compare: what was measured, who was sampled, which year, and which clock the hours use. Neither measures your audience; each pools other people's channels.
What YouTube says about publish time
YouTube's performance FAQ is the one source that describes how the system treats publish time. This table sets its statements beside what they mean for a Shorts schedule.
| What YouTube says | What it means for you |
|---|---|
| Publish time is not known to impact a video's long-term performance | An off-peak time is not known to determine long-term performance; early views may still differ |
| The recommendation system aims to deliver the right videos to the right viewers regardless of when a video was uploaded | The source does not establish that timing has no effect; viewer interest is one signal |
| Publish time is important for formats such as Live and Premiere | Timing matters for those formats; a Short is neither |
| Publishing when your audience is most active can be beneficial for early viewership | A good window can lift the first hours, which is what your test should measure |
Buffer's data leads it to a similar view: it describes Shorts as continuously resurfaced by YouTube's recommendation engine over days or even weeks, so the first upload hour counts for less. Buffer also finds that long-form videos and Shorts peak at different hours, mornings for long-form and evenings for Shorts, so a schedule built for one format is not a schedule for the other. For the format itself, see our short-form video best practices.
Open your own viewer-activity report first
YouTube points creators to one report for this question. In YouTube Studio, open Analytics, then Audience, and find When your viewers are on YouTube, which charts when your viewers are online across all of YouTube over the past 28 days. YouTube's tips page says this data can help you determine what time and day is best to upload.
- It covers all of YouTube. The help page describes your audience's activity across your channel and all of YouTube, not only Shorts, so treat it as context for the test and not as a Shorts verdict.
- It looks back 28 days. Check it again after a month, because your audience changes.
- A new channel has little to show. Bytecap warns that a young channel's report can be too sparse to use; if the report is unavailable and you need starting windows, tentatively try Friday and Saturday at 4 p.m. in your intended audience's time zone. This is an editorial assumption for two distinct slots; the studies do not establish whose local clock applies.
- Check the clock before you copy an hour. Buffer describes the chart as set in local time; YouTube's help page does not identify the clock, so confirm the setting shown in Studio before scheduling. If the chart is in your Pacific clock and your viewers are on US Eastern time, a 4 p.m. peak on the chart is 7 p.m. for them, and you schedule it for 4 p.m. Pacific with no further conversion.
Pick two test windows
Choose two windows and nothing more. More windows means too few Shorts in each to read. This table lists the candidates, with the source for each.
| Window | Where it comes from | Use it when |
|---|---|---|
| The peak in your own report | YouTube Studio, last 28 days | You have enough audience data |
| Friday 4 to 7 p.m. | Buffer's top three Shorts slots | You have no audience data yet, or your viewers skew evening |
| Saturday at 4 p.m. | Adobe's best single day-and-hour slot | You want the second study's answer in the test |
| Any hour on Tuesday | Adobe's best day for overall views and likes | You want to test Adobe's day without its hour |
| An off-peak hour, such as midday or late evening | Adobe: posts outside 8 to 10 a.m. did better than those inside it | Your niche posts heavily in the morning |
If your report is usable, use its peak for A and a different study slot for B. For example, if the peak is Friday at 4 p.m. in your chosen clock and overlaps the Buffer candidate, try Saturday at 4 p.m. as B. Without a usable report, Friday and Saturday at 4 p.m. in your intended audience's time zone are an illustrative pair. If you do not know an intended audience time zone, choose and record one working time zone. This clock choice is an editorial assumption, not a verified interpretation of either study or a promise that either slot is best. If you post long-form videos too, keep them out of the test: Buffer's evidence puts long-form at its best in the morning, with almost no overlap with Shorts evenings.
Two frames from a licensed stock clip, approximately 0.4 and 3 seconds in. The first shows the cup before the pour; the second shows the stream. If clips in two posting windows open with different action, that difference can also affect a comparison.
Publish comparable Shorts in both windows
A useful timing comparison tries to keep format and audience similar, but different videos and time slots can still confound the result. YouTube advises comparing videos of the same format, since audience behavior differs across formats. So use the same kind of Short and a similar length in both windows, and put one Short in each window every week, so a busy week or a quiet week hits both windows alike.
- Make ten Shorts in one format. Five per window is our small planning heuristic, not a statistically established minimum. A median is usually less sensitive to extreme values than a mean, but even one changed observation can change it. Treat the result as a pointer, not proof.
- Schedule each one in YouTube Studio. The upload flow has a Schedule option: choose the date and the video stays private until then. Studio costs nothing, which is all this test needs; a separate scheduler only earns its place if the same clip also goes to other apps.
- Post one Short in each window every week for five weeks. Swap which topic goes to which window, so the stronger ideas do not all land on one side.
- Write the window and your viewers' local time in the test log before you publish, so the result cannot be bent afterward.
Three mistakes spoil most timing tests, and each has a simple fix.
- Changing two things at once. A new hook style in window B makes the hour impossible to read. Keep format, length and topic type the same on both sides.
- Reading one lucky Short. A single clip that takes off says nothing about the hour. Compare medians, while remembering that one observation can still change a median.
- Testing through a launch or a holiday. A sale or a news spike lifts one week and not the next. Pause the test for that week and add a week at the end.
Posting every day is not required for this. Our guide to posting cadence covers pace, and our content calendar guide shows where to record each window; the test only needs the same pace in both windows.
One comparable Short goes in each window every week for five weeks, with the topics swapping sides. Original diagram of this guide's example plan, drawn from the worked example below.
Read the first 24 hours and keep the winner
YouTube Analytics has what the test needs. Its Advanced mode lists special reports including first 24 hours of video performance, lets you add a comparison of videos, groups or time periods, and lets a group hold up to 500 items. Put the Shorts from window A in one group and window B in another.
Read engaged views, not raw views. YouTube counts a Shorts view each time a Short starts to play or replays, with no minimum watch time, and keeps a second measure called Engaged views, which shows how many viewers chose to continue watching. YouTube says you can use Engaged views to compare performance across all of your Shorts.
| Read | Rule (ours, YouTube publishes none) | What it tells you |
|---|---|---|
| Median engaged views in the first 24 hours, window A against window B | With five valid observations in each window and a positive comparison median, consider a window ahead when its median is at least 1.5 times the other's. Five Shorts can differ on topic alone. | Whether early-view medians differ; the test does not prove why |
| Engaged views at 28 days, same Shorts | Keep the window only if it is still at least 1.5 times ahead | YouTube says publish time is not known to affect long-term performance; a fading gap is descriptive, not proof of cause |
| Gap under 1.5 times | Treat it as inconclusive and publish when it suits you | This small sample cannot establish whether timing affected results; continue testing topics and hooks |
| Zero, missing or fewer than five valid observations per window | Treat the comparison as inconclusive; do not divide by zero | There is not enough usable data to apply this heuristic |
A small or unclear gap is inconclusive; it does not establish that timing has no effect. Choose a practical schedule and continue learning.
A worked example: one five-week test, start to finish
Here is the whole test for an imagined founder of a small skincare brand. She works in San Francisco, and her viewers are mostly on US Eastern time. The numbers below are invented to show the arithmetic. They are not a measured result from any channel.
- Window A from her own report. In Studio she confirms that When your viewers are on YouTube is shown in her own Pacific clock. Its busiest weekday band is around 4 p.m. Pacific, which is 7 p.m. for her Eastern viewers, so window A is Wednesday at 4 p.m. Pacific (7 p.m. Eastern). She schedules it as the chart shows it, with no conversion.
- Window B from a study. She takes Buffer's Friday 4 p.m. slot. Buffer says its hours apply "in your local time" but not whether that means hers or her viewers', so she uses her viewers' time as an unverified test assumption: 4 p.m. Eastern, which she schedules for 1 p.m. Pacific.
- Ten comparable Shorts. Each one is a 25-second tip with the product on screen and the same caption style. One goes out in each window every week for five weeks, and she swaps which topic goes where.
- The 24-hour read. The five Shorts in window A had 620, 410, 380, 900 and 450 engaged views after 24 hours, a median of 450. Window B's had 240, 300, 260, 510 and 280, a median of 280. A's median is 1.6 times B's, so A clears the 1.5 times bar for early views.
- The 28-day read. Once the last Short is 28 days old, window A's median is 1,700 engaged views and window B's is 1,550. That is a gap of about 1.1 times, below the 1.5 times bar, so the early lead did not hold. With five Shorts a side, she reads this as a pointer, not proof.
Her verdict is inconclusive. Her Wednesday Shorts had a higher first-day median in this fictional sample, but the two windows were close after four weeks. With only five Shorts per window, the difference does not establish that timing caused either result. She keeps Wednesday at 4 p.m. Pacific because it suits her week and continues testing topics and hooks.
Invented numbers show the arithmetic: the early read clears the 1.5 times bar and the 28-day read does not, so the comparison ends inconclusive. Original diagram of the worked example's figures, November 11, 2026.
Where Superdirector fits
Superdirector publishes this guide. Its $199/month Growth offer is for one AI-run social account in your brand's direction. For supported channels, current publishing accepts eligible Instagram carousel images and one link-free X text post under the account's standing posting permission; confirm access and readiness for your chosen account. It does not post YouTube Shorts today. AI video and fully autonomous channels without a person remain upcoming. The $29/month Starter plan supports hands-on work from Claude or ChatGPT. YouTube Studio is enough for the test above. See current plans.

