YouTube Shorts Algorithm Explained + Tips to Grow in 2026
Gretchen Oestreicher
05 August 2026 
More than 60% of every organic view on YouTube now comes from the Shorts feed. That makes the Shorts algorithm the main way people find video on the platform. If you post Shorts, it’s the system deciding who sees your work, so it’s worth knowing how it thinks.
We pulled the numbers from Metricool’s 2026 YouTube Study, which looked at 799,718 videos across 71,177 accounts. Two figures stand out: views per Short more than doubled, and average watch time dropped to about 16 seconds. Together they say a lot about what the algorithm rewards right now.
What Is the YouTube Shorts Algorithm?
The YouTube Shorts algorithm is the system that decides which Shorts show up in someone’s feed, and in what order. It watches what each person plays, skips, and enjoys, then builds a feed around that. Most of what people see comes from channels they don’t follow yet.
It works on its own, separate from the system behind long-form video. Your Short competes against other Shorts for a spot in the feed. What decides that spot is how people react to it, not how big your channel is. YouTube puts it plainly in its own docs: “Our system has no opinion about what type of Shorts you make.”
That matters, because of where views actually come from. In our sample, the Shorts feed drove more than six in ten organic views. Subscribers and YouTube search came next, and both trailed a long way behind.
| Source | Views | % |
|---|---|---|
| Shorts feed | 19,895,798,605 | 61.54% |
| Subscribers | 3,656,802,001 | 11.31% |
| YouTube search | 2,845,787,066 | 8.80% |
| Related video | 1,866,928,375 | 5.77% |
| Playlist | 1,633,177,994 | 5.05% |
| Channel page | 1,007,592,625 | 3.12% |
| Hashtags | 9,036,968 | 0.03% |
YouTube traffic sources (Metricool 2026 YouTube Study)
Almost no one lands on a channel on purpose to look for videos. Nearly everything else runs through recommendation. So it pays to know how that recommendation works.
How the YouTube Shorts Algorithm Works in 2026
YouTube actually tells you which signals its Shorts system uses. The confirmed list is shorter than most algorithm guides make out.
- Viewed vs. Swiped Away: when your Short shows up in someone’s feed, this is how often they stop to watch instead of scrolling past. It’s the first test every Short has to pass, and you can track it in YouTube Studio next to the Shown in Feed metric.
- Average View Duration and Average Percentage Viewed: how long people stay, and how much of the Short they finish. A 20-second Short watched to the end beats a 3-minute Short people quit at second eight.
- Satisfaction: likes, dislikes, and those post-watch surveys. YouTube asks people what they thought and feeds it back in. A Short can rack up watch time and still lose its reach if people didn’t enjoy it.
- Watch History and Topic Interest: what someone has watched before shapes what they see next. How much people care about a topic affects how far a Short on that topic travels.
- Sounds: using a trending or sampled audio track is a signal too, and it can put your Short on that sound’s page.
- Competition and Seasonality: how your Short does against everything else posted on similar topics at the same time, and how interest in a topic rises and falls through the year.
What Isn’t a Ranking Signal
Three things get repeated a lot, and none of them show up in YouTube’s docs. Building a plan around them is wasted effort.
Sharing your Short to TikTok, Instagram, or Facebook won’t win it any favor with YouTube’s algorithm. Cross-posting can bring in outside traffic, and external links drove a small slice of views in our sample. But that traffic isn’t something the ranking system counts.
Your subscriber count doesn’t set your reach. The feed is built to show people channels they don’t follow yet, and the data backs that up. That’s good news if you’re just starting out.
How often you post isn’t a ranking factor either. YouTube says there’s no minimum frequency you need to hit. Posting more just gives you more chances to show up. It doesn’t change how any single Short is judged.
Views vs. Engaged Views: Read Your Analytics Correctly
Since 31 March 2025, YouTube counts a view every time a Short starts or replays, with no minimum watch time. That matches how TikTok and Reels count views. It also made the big number on your dashboard jump.
The stricter original count still exists, now called engaged views. You’ll find it in Advanced Mode in YouTube Studio. These are the views that count for money. They feed revenue sharing, and the Shorts path into the YouTube Partner Program needs 10 million of them in 90 days.
Two things follow from this. First, your view count is now a reach number, so check it alongside viewed vs. swiped away and average view duration before you call a Short a hit or a miss. Second, any comparison across March 2025 straddles a change in what counts as a view. That’s part of why engagement rates look lower everywhere in 2026.
Shorts Performance in 2026: What the Data Shows
Brands posted more Shorts this year, week after week. Views went up with them. Watch time went the other way.
| 2025 | 2026 | Difference | |
|---|---|---|---|
| Views | 2,343.11 | 5,320.59 | 127.07% |
| Estimated minutes watched | 1,860.19 | 1,375.58 | -26.05% |
| View duration (minutes) | 0.79 | 0.26 | -67.43% |
| Engagement | 4.94% | 1.85% | -62.61% |
| Interactions | 115.80 | 98.32 | -15.10% |
| Posts per week | 2.26 | 2.72 | 20.38% |
Shorts performance on YouTube, 2025 vs. 2026 (Metricool 2026 YouTube Study)
Each Short now gets watched for about 16 seconds, roughly half a minute less than a year ago. More Shorts are reaching more feeds than before. The catch is that people swipe away sooner and interact less once they get there.
Being Small Is an Advantage Right Now
The smallest accounts are growing fastest in the whole sample. Channels with a few thousand subscribers more than tripled their views and saw interactions climb too. The biggest channels, those over a million subscribers, did the opposite and lost most of their views per Short.
| Segment | Views 2025 | Views 2026 | Views change | Interactions change |
|---|---|---|---|---|
| Tiny (<2K) | 481.53 | 1,289.72 | 167.84% | 15.29% |
| Small (2K-10K) | 1,760.39 | 5,466.07 | 210.50% | 41.47% |
| Medium (10K-100K) | 5,277.24 | 13,051.93 | 147.32% | 4.68% |
| Big (100K-1M) | 15,917.15 | 35,421.36 | 122.54% | -21.56% |
| Huge (1M+) | 46,326.11 | 11,953.86 | -74.20% | -41.25% |
Shorts performance by account size (Metricool 2026 YouTube Study). Raw interaction counts omitted for width. The full grid is in the source study.
This fits what YouTube tells us about the algorithm. A Short rises or falls on how people react to it, so a video that hits with a small audience can travel to a much bigger one. Those same small channels were also the most likely to move up a subscriber tier this year, well ahead of the YouTube average.
How Shorts Affect Your Long-Form Video
Channels that post no Shorts see below-average views on their long-form videos. Channels posting a handful of Shorts a month do meaningfully better, roughly 50% above the study-wide average.
| Shorts segment | Shorts posts | Views | Difference |
|---|---|---|---|
| No Shorts | 0 | 4,364.72 | -27.08% |
| Below-average Shorts volume | 4.91 | 8,789.87 | 46.86% |
| Above-average Shorts volume | 35.04 | 4,434.79 | -25.90% |
How posting Shorts affects long-form video performance (Metricool 2026 YouTube Study)
It has a limit, though. Push the volume high enough and the lift disappears. Post Shorts constantly and your long-form views slide back to about where the no-Shorts channels sit.
This isn’t proof that Shorts cause the lift. Channels that post Shorts might just be more active, or more tuned in to their audience. One thing is clear, though: Shorts don’t cost you anything on the long-form side. YouTube says Shorts performance can’t hurt your long-form recommendations, and every video is judged on its own. If you’re posting no Shorts right now, it’s worth a test.
How Long a Short Keeps Working
A Short does most of its work early. The first day brings a big share of its views and roughly a third of its interactions. The first three days are the strongest stretch, while the algorithm tries it out on your subscribers and on people with matching interests. Most of a Short’s interactions come in during the first week and a half.
After that, it slows down without fully stopping. YouTube keeps showing your Short to new audiences, so one with a solid topic can keep picking up views for months at a quieter pace. That kind of shelf life is rare in short-form video. It rewards Shorts built around topics that still matter after the trend fades.
YouTube Shorts Algorithm Tips to Grow
Earn the View Before You Earn the Watch Time
Every Short faces the same first test. Did the person stop, or did they keep scrolling? Your viewed vs. swiped away rate in YouTube Studio shows how often you pass it. Fix this before anything else, because nothing happens until you do.
Give people a reason to stay in the first second. Show the payoff early, lead with your best moment, and cut anything that stalls. With watch time sitting at 16 seconds, a Short that needs 10 seconds to get going has spent most of its budget on the intro.
Post 2 to 4 Times a Week, Then Scale if You Can Sustain It
Per post, 2 to 4 Shorts a week is the range where each video does best. Post more and the per-post numbers dip a little, though never enough to make posting more a bad call.
| Frequency | Views | Interactions |
|---|---|---|
| ≤0.5/week | 3,535.34 | 60.78 |
| 0.5-1/week | 5,811.71 | 112.19 |
| 1-2/week | 5,157.37 | 91.53 |
| 2-4/week | 5,872.34 | 130.43 |
| 4-7/week | 5,086.81 | 97.59 |
| 7+/week | 5,077.65 | 95.82 |
Performance of YouTube Shorts by weekly posting frequency (Metricool 2026 YouTube Study)
Add it up over a month, though, and there’s no clear ceiling. Post several times a week and your monthly views climb well past what the 2-to-4 group pulls in. An endless feed gives every upload another shot to show up. Volume won’t win you special treatment, so make it a production call: post as often as you can without letting the quality slip.
Use Hashtags to Add Context
Shorts with hashtags got fewer views but higher engagement than Shorts without. Across the whole sample, hashtags drove under 1% of views. That tells you most of what you need to know about them as a discovery tool.
They still earn their place. They tell the algorithm what your Short is about and tie it to trends and niche communities. The people who find you through them tend to care more about the topic, too. Stick to 1 to 5 that actually describe the content.
Make Shorts People Are Glad They Watched
Satisfaction sits right next to watch time in YouTube’s confirmed signals. It’s measured through likes, dislikes, and post-watch surveys. A Short can hold attention and still leave people feeling like they wasted their time. The system is built to spot the difference.
So make Shorts that would still be worth watching six months from now, and that give people a reason to check out the rest of your channel. That’s what turns a spot in the feed into a new subscriber.
Post With the Crowd or Against It
Most Shorts in the study went up between 2 and 6 p.m. Post in that window if you want to ride peak activity. Pick a quieter time if you’d rather face less competition for the same feeds.
What’s New for Shorts in 2026
Length
Shorts can now run up to three minutes, a change that arrived in October 2024 for square and vertical uploads. Most licensed music still works for up to 90 seconds inside a three-minute Short. A longer runtime doesn’t change what the algorithm measures. And with watch time at 16 seconds, a longer Short needs a real reason to be longer.
AI Creation Tools
Veo 3 Fast brings text-to-video with sound right into the Shorts camera, free, starting in the US, UK, Canada, Australia, and New Zealand. Edit with AI builds a first draft from your camera roll, with transitions, music, and voiceover. Reimagine turns a single frame into a short generated clip. Ingredients to Video makes a vertical clip from up to three images you upload. A new Gemini-powered remix feature drops you or an object into another creator’s eligible Short. It started rolling out free in May 2026. As of December 2025, more than a million channels were using YouTube’s AI tools every day.
Policy on AI and Repetitive Content
YouTube updated its inauthentic content policy in July 2025. Mass-produced, repetitive content can’t be monetized anymore: templated formats, identical slideshows, the same voiceover on every upload. AI-assisted content is still fine when it adds real human value, and each video has to be meaningfully different from the last. Using AI tools is fine, as long as you’re not churning out near-identical copies.
Monetization
Shorts revenue comes from a shared pool, and monetizing creators keep 45% of their cut. Engaged views are what feed both Partner Program eligibility and revenue, not the headline count. Reported Shorts RPM swings a lot with audience location, topic, and licensed music, so your own figure in YouTube Studio beats any benchmark you’ll read online.
What the Algorithm Actually Rewards
The Shorts algorithm decides most of what gets seen on YouTube. What it responds to is narrower than the myths suggest: did people choose to watch, how long did they stay, how much did they finish, and were they glad they did.
The 2026 data points the same way throughout. Views per Short have more than doubled, so reach is easier than it used to be. Watch time is down to about 16 seconds, so holding attention is the harder job now. Smaller accounts are growing fastest. The system cares about how people react to a Short, not how long you’ve been at it, so anyone can compete.
Connect your YouTube channel to Metricool and see your Shorts and long-form videos side by side. Track engaged views and watch time as you post, and spot which content is actually holding attention.
You can also compare your own numbers against the 71,177 accounts in this study with Metricool’s Claude benchmarking skill. It gives you back a scored report by subscriber tier and format.
