The best time to post YouTube Shorts, from 14,122 Shorts

Every guide to Shorts recommends a different time. One says Friday at 4pm, another says weekday lunchtimes, another says weekday afternoons in your local time. That makes planning harder if you manage several channels or clients. Most guides rank raw engagement, so a few huge channels posting together can make an hour look better.

I compared 14,122 public Shorts from 227 channels across 60 niches with each channel's usual views. A channel averaging ten million views counts the same as one averaging five thousand. A Short scores high only if it beats its channel's usual views. With this method, the hour barely matters, and Saturday is the only day that does consistently better.

At a glance

The hour you upload barely matters

No hour gave Shorts a meaningful advantage. After smoothing across neighbouring hours, views stay within about 4% of the average Short. The six four-hour blocks stay within about 2%.

About half the Shorts in each block beat their channel's usual views, from the middle of the night to the evening. If a block helped Shorts get more views, more than half would beat their channel's usual count.

This sample can't rule out a real difference of a few percent. But that's too small to change your schedule for. Upload when the Short is ready.

Saturday leans ahead, and Monday and Tuesday trail

Saturday was the only day Shorts did consistently better, about 6% above the rest of the week. Sunday was next. Monday and Tuesday did worst, about 4% below the other five days combined.

I checked whether the Saturday result held in different groups:

• Without the viral outliers: the difference was about the same. Shorts with unusually high views for their channel don't explain it.
• Channels found through niche searches only: the difference was about the same.
• Channels whose typical Short gets 10,000 to 100,000 views: Saturday's advantage was largest here, about 10%.
• Channels whose typical Short gets over a million views: Saturday had no advantage, though this group has only 16 channels.

Unlike LinkedIn users, these channels don't post less on weekends. Uploads were about evenly spread across the week, so Saturday Shorts don't face less competition. More people may watch on their day off, or channels may save better Shorts for weekends. This data can't tell which.

If you make Shorts in batches, post one of the week's better ones on Saturday and check whether your channel does better too.

Friday evening, the slot most guides name, didn't stand out

Shorts uploaded on Friday between 4pm and midnight ET did about the same as those uploaded at other times. Buffer ranks that time first for Shorts, and SocialPilot ranks Friday 4pm to 7pm highest. Here, views were slightly below average, but the difference was too small to matter.

If you're choosing between Saturday and Friday, Saturday did about 6% better.

Scheduling doesn't help or hurt

About 1 in 6 Shorts were uploaded within the first minute of an hour. People rarely press publish at exactly 3:00, so most were scheduled uploads. They did about as well as Shorts uploaded at other minutes.

Scheduling is convenient. You can post on Saturday without working that day.

How this compares to what's already published

• Buffer studied 1.8 million YouTube videos, with a separate Shorts section. It recommends Friday at 4pm, ranks Friday, Saturday and Thursday as the best days, evenings as the best times, and Tuesday and Monday as the worst days.
• SocialPilot studied about 301,000 videos from 27,000 channels using its scheduler. It recommends 12pm to 2pm or 6pm to 7pm local time for Shorts, ranks Friday, Saturday and Thursday as the best days, and Friday 4pm to 7pm as the best time.

These results agree on the days: Saturday near the top, Monday and Tuesday near the bottom. They don't support the recommended hours. Neither guide gives a margin of error for its times.

The studies differ because they ask different questions. Adding up engagement across millions of videos finds when the most people watch and channels post. This study asks whether a Short beat its channel's usual views. That's more useful to you, since you can't change who else posts at 4pm on a Friday.

How I measured it

The sample has 14,122 Shorts from 227 channels, uploaded between two weeks and a year before collection. Four YouTube endpoints on AnyAPI returned them in 876 calls for $1.00.

1. Find channels that post Shorts. One youtube.search_shorts search per niche covered 60 niches, from cooking and fitness to woodworking, nursing and crypto, filtered to the last month's Shorts. One youtube.video lookup on up to seven results per niche identified each Short's channel. One youtube.trending_shorts call added that day's trending channels, for 263 channels total.
2. Get each channel's Shorts. Request up to two pages of youtube.channel_shorts per channel, newest first, with 48 Shorts per page, upload times and view counts.
3. Find missing times. I looked up 90 Shorts with missing upload times using youtube.video. That found times for 64; I dropped the rest.
4. Allow time for views. I dropped 5,896 Shorts younger than two weeks, 969 older than a year, and channels with fewer than five Shorts left. That left 227 channels and 14,122 Shorts.
5. Score each Short against its channel. Divide its views by its channel's median views. Cap scores at 6x so one viral Short counts once. 1,403 Shorts hit the cap, about one in ten, far more than in the LinkedIn study. Shorts' views vary widely.
6. Group and average. Convert upload times to US Eastern. Group scores by weekday, hour and four-hour block, then take each group's geometric mean. Resample every comparison by channel 1,000 times to check how much a different set of channels could change it.

This call gets one channel's Shorts. The response shows only the fields used:

curl -sS -X POST https://api.getanyapi.com/v1/run/youtube.channel_shorts \
  -H "Authorization: Bearer $ANYAPI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{"handle":"@MrBeast"}'

{"output":{"data":{"nextCursor":"cur_...","shorts":[{"id":"41UzbuW6wJs","title":"Can A Robot Do My Chores?","createdUtc":1793379604,"views":49424472,"viewsAvailable":true},"..."]}},"items":48,"costUsd":0.0012}

The channel listing returns exact view counts. YouTube's Shorts shelf and search results show rounded figures (8.9M, 875K).

The limits of this data

• Total views, not first-day views. Views were counted two weeks to a year after upload, so they can't show how a Short did in its first 24 hours.
• US Eastern for everyone. The listings don't give channel locations. All upload times use Eastern, so results for a channel in London or Mumbai shift by its time difference. Days change less, but a late-night upload abroad can fall on a different Eastern day.
• Views only. The study doesn't measure likes-per-view, watch time or subscribers gained.
• Better Shorts may go out on better days. If channels save their best Shorts for Saturday, those Shorts could explain the results, rather than the day.
• Recent Shorts weigh more. Two pages contain about 96 Shorts, so daily uploaders contribute only their last few months.
• A searched sample. The channels came from 60 English niche searches and one trending list, favouring active channels in those niches. Different channels could change a result this small.
• Few very large channels. Only 16 channels typically get over a million views per Short, too few to draw conclusions about that group.

Saturday did about 6% better. Different channels could put that difference anywhere from about 1% to about 11%. Differences between hours are smaller.

What to do with it

1. Upload when it suits you. No hour gave Shorts a lasting advantage.
2. Post one of your better Shorts on Saturday. It was the only day that did consistently better, by about 6%.
3. Don't plan your week around Friday evening. The time most guides recommend did about average here.
4. Post less on Monday and Tuesday if you can. They did slightly worse, matching Buffer's worst days.
5. Use scheduling freely. Scheduled Shorts did as well as manually published ones.
6. Check your own channel. Get your Shorts with youtube.channel_shorts, drop those younger than two weeks, divide each Short's views by your median, and compare Saturday with the rest. One page of 48 Shorts costs $0.0012.

The TikTok study, Instagram study and LinkedIn study use the same method. Use the YouTube endpoints to repeat this for your channel or a client's.

Use data responsibly and follow AnyAPI's Acceptable Use Policy.