Why Your YouTube Shorts Are Not Getting Views: The Five Gates Between Upload and a Full Feed

The complaint is always the same shape. Twelve Shorts, each one better edited than the last, each one sitting somewhere between 40 and 300 views, while the channel's long-form videos tick along normally. Nothing was struck, nothing was warned, and Studio shows a flat line that looks less like a bad video than an unplugged cable.

Most advice answers this with content notes — hook harder, cut faster, post more. Sometimes that is the answer. Often it is not, because the Short never got as far as being judged on its content. Between pressing publish and a full feed there are several separate systems that can each stop a Short dead, and they fail in ways that look identical from the outside: a small number that stops moving.

So this is a diagnostic rather than a pep talk. What the view number actually means now, the five gates a Short has to pass, how to tell from Studio which one you are stuck at, and which of the popular explanations are not mechanisms that exist. Along the way, the part that is genuinely a packaging problem — and it is a bigger part than most creators expect.

Start by checking which number you are looking at

Shorts changed their unit of measurement before the rest of YouTube did. From 31 March 2025, YouTube counts a Shorts view the moment the video starts to play or replay, with no minimum duration. The older definition — someone stayed past the first couple of seconds — did not disappear. It was renamed engaged views, and YouTube has been clear that engaged views, not views, are what feed revenue sharing and the Shorts route into the Partner Programme.

Long-form caught up on 24 August 2026, when the public view count moved to first-frame counting across videos, Shorts and live streams alike. If you are comparing this month's Shorts against something from early 2025, you are comparing two different instruments; the views drop diagnostic covers that discontinuity in detail.

The practical consequence is that a Short can have two very different failures and only one number to show for them. Four thousand views with 900 engaged views is a Short that is being distributed and rejected: people are being shown it, it starts playing, and they are gone before the count of people who stayed can rise. Four hundred views with 380 engaged views is almost the opposite — nearly everyone who saw it stayed, and the problem is that almost nobody saw it. The first is a packaging problem. The second is a distribution problem, and no amount of re-editing the hook will touch it.

Open Advanced mode in Studio, switch the metric to engaged views, and write both numbers down for your last ten Shorts before you change anything. Nearly every wrong decision in this area comes from skipping that step.

The five gates between upload and a full feed

It helps to stop thinking of the Shorts feed as a single judgement and start thinking of it as a sequence. Each gate has a different failure signature in Studio, and the fixes have nothing in common with each other.

Gate	What it decides	What failure looks like in Studio
1. Classification	Whether the upload is a Short at all	No Shorts feed traffic source; the video sits with your long-form
2. Eligibility	Whether it may be recommended	Shown in feed barely moves from the first hour, on every upload alike
3. The first test	Whether a sample audience stops	Shown in feed rises, swiped away dominates, views flatten early
4. Retention	Whether the people who stopped stayed	Healthy swipe-through, weak average percentage viewed, no replays
5. Expansion	Whether it travels past the first audience	Good numbers everywhere, distribution stops at a ceiling

Work them in order. Gate three is where most creators start, and gates one and two are where a surprising share of the dead Shorts actually are.

Gate one: YouTube has to classify it as a Short at all

Since 15 October 2024, classification is automatic and mechanical. An upload that is square or taller — an aspect ratio from 1:1 up to 9:16 — and three minutes or shorter is treated as a Short. Nothing else is required. The #Shorts tag that everyone still types into titles has not been necessary for years, and adding it to a horizontal eight-minute video does nothing at all.

Which means the two ways to fail here are boring and common. A cut that runs to 3:04 because the outro drifted is not a Short; it is a very short vertical video, which lands in a completely different competition with no swipe feed behind it. And an export that pads a vertical clip into a 16:9 frame with black bars is, as far as classification is concerned, a horizontal video with a small picture in the middle of it.

The check takes ten seconds: open the video in Studio and look for the Shorts icon on the entry, then open its analytics and look for Shorts feed in the traffic sources. If Shorts feed is absent on a video you intended as a Short, you have found your answer and it has nothing to do with content. Trim the duration or fix the export ratio and reupload; classification is set at upload and does not change later.

Gate two: the settings that quietly delete your distribution

Shorts distribution is overwhelmingly personalised recommendation. Search, your channel's Shorts grid and the subscription shelves do deliver views, and they keep delivering them for months, but on most channels they are a slice rather than the body of it. The feed is the body of it, and it is built entirely out of personalised recommendation — so any setting that pulls a video off personalised surfaces removes the mechanism that was going to deliver the views.

The made-for-kids setting is the most destructive of these, and the one most often applied by accident — sometimes at channel level, where it then applies to everything you upload. YouTube's own documentation is explicit that content set as made for kids loses a long list of features, including personalised advertising, comments and notifications. The blast radius is wider than the feature list suggests, because those features are downstream of the same personalisation the feed runs on. If your Shorts are dying uniformly at zero, check the audience setting on the channel first and the video second; our guide to what the made-for-kids setting disables has the full list.

Age restriction has a similar effect for a different reason: restricted content is hidden from signed-out viewers and pulled from most recommendation surfaces, which is most of what a Short has. A copyright claim can restrict a video in specific countries, and if the restricted countries happen to be where your audience is, the feed has nobody to show it to. None of these produce a warning that looks like a problem — the video is public, the Short plays when you open the link, and only the distribution is missing.

One thing that is not on this list: limited or no ads. The yellow icon changes what a video earns, not how far it travels. If you have been treating a yellow icon as an explanation for low views, the limited ads explainer is the short version of why those are separate systems.

Gate three: the first test, and the two numbers that describe it

A published Short gets shown to a small sample of the feed. Studio reports this as two things side by side: how many times the Short was shown in the feed, and, of those, how often viewers watched rather than swiped away. That pair is the first real judgement, and it happens in minutes, not days.

It is worth being honest about the benchmark question, because a great deal of confident numbering circulates here. YouTube publishes no target swipe-away rate and no threshold for "good". The specific percentages you find quoted in creator guides are third-party estimates from third-party datasets, and they are not what your feed is graded against. What is genuinely useful is the internal comparison: your last ten Shorts, same rough format and length, ranked by swipe-away. The three worst have something in common, and it is usually visible in the first frame.

The failure signature here is distinctive. Shown-in-feed climbs steadily for an hour or two, views crawl, and then both flatten. That is the system doing exactly what it is designed to do: it offered your Short to people, they declined, and it stopped offering. Nothing was suppressed. A test was run and lost.

The swipe is a packaging problem wearing different clothes

On long-form, the thumbnail and title do the selling and the video does the delivering. On Shorts, both jobs collapse into the same few hundred milliseconds, which is why so many creators who are good at video are bad at Shorts. The first frame is a thumbnail that cannot be chosen separately from the footage, shown to someone whose thumb is already in motion.

Four things decide it, in roughly this order of impact:

• What is legible at a glance. A wide establishing shot reads as nothing at phone size. A face, a single object, or four words of large text reads instantly. This is the same constraint that governs long-form thumbnails, and the same composition rules transfer almost unchanged.
• Whether anything moves in the first half-second. A static opening frame with a voice fading in is the single most common swipe cause. Movement in the frame buys the sentence that follows.
• What the audio does at zero. Silence, a breath, or a three-word intro all read as "this has not started yet", and the feed does not wait for things to start.
• Whether the promise arrives before the setup. The hook rules for long-form apply here compressed by a factor of ten: state the thing, then explain it, never the reverse.

Since 25 July 2026, Shorts have also supported custom covers, starting with creators in the Partner Programme. A cover controls what the Short looks like in search, on your channel's Shorts grid and on the shelves where it is browsed rather than swiped — it does not replace the first frame in the feed itself, which is where most Shorts views come from. The Shorts cover guide covers where each one is actually seen, which is the detail that decides whether making covers is worth your time.

Gate four: staying, finishing, looping

Passing the swipe test buys you a few seconds of attention and a second judgement. Studio reports average percentage viewed for Shorts, and because Shorts loop, that number can exceed 100% — a 15-second Short at 140% is being watched roughly one and a half times per viewer. That is the strongest signal a Short can produce, and it is far easier to engineer than most creators assume.

Length discipline does most of the work. The same content at 22 seconds and 55 seconds will produce very different percentages, and the shorter version is not merely better scored — it is better, because the padding that fills the extra 33 seconds is what people are leaving during. Cut every clause that is not load-bearing, then cut the first sentence, which is almost always a throat-clear.

The second lever is the ending. A Short that stops on a beat rather than trailing into an outro tends to loop, because the viewer's brain reaches for the thing they missed. A Short that ends with "so if you enjoyed that, subscribe" reliably does not, because the request marks the content as finished. Keep the call to action, move it into the description or a pinned comment, and let the last frame carry the loop.

Retention failure has a clean signature: healthy shown-in-feed, acceptable swipe-through, then views that stop climbing anyway. The feed took a look, the sample stayed long enough to be counted, and the retention was not good enough to justify a wider push.

Gate five: expansion past the first audience

The last gate is the one creators misread most often, because everything looks fine and the numbers stop anyway. A Short that performs well with the audience it was first shown to still has to perform with the next audience out, who share fewer interests with your existing viewers. Each expansion is a fresh test against a less sympathetic sample.

This is where a channel's topical coherence starts to matter. YouTube evaluates each piece of content individually — mixing Shorts, long-form and livestreams does not confuse anything — but the system's understanding of who enjoys your work is built from what your work has been. A Short that is unlike everything else you have published has no established audience to be matched against, so it gets tested against a guess, and guesses convert worse.

A ceiling that reproduces across many Shorts at roughly the same number is usually this: the content is good enough for your existing audience and not distinctive enough to survive the jump to the next one. The fix is not more uploads. It is a sharper premise — a Short that someone with no prior interest in your channel would still stop for.

Reused, watermarked and mass-produced: what is documented and what is folklore

This area attracts more confident misinformation than any other, so it is worth separating what YouTube has written down from what creators have inferred.

Documented: YouTube's Shorts monetisation policies state that some views are ineligible for revenue sharing, and give examples — non-original Shorts such as unedited clips from someone else's films or television, reuploads of other creators' content, artificial traffic from bots, and views of content that does not meet advertiser-friendly guidelines. In July 2025 YouTube renamed its repetitious content policy to inauthentic content, which tightened the language around mass-produced and templated uploads. Those are monetisation rules with published text behind them.

Not documented: a reach penalty for a visible watermark from another platform. The claim is everywhere, and the underlying instinct is sound — YouTube has added its own watermark to Shorts downloaded for cross-posting since 2022, so the platforms clearly care about provenance — but YouTube has not published a statement that a TikTok watermark reduces distribution. Treat the clean-export habit as cheap insurance rather than as a known mechanism, and do not let it become the explanation for a Short that failed for ordinary reasons.

Three 2026 changes that moved the target

The feed you are publishing into is not the feed from a year ago, and three changes are worth factoring into a diagnosis.

Shorts moved into the living room. At its Made on YouTube event in September 2026, YouTube said viewers are watching more than two billion hours of Shorts on televisions each month. A vertical video on a large, distant screen is a different object: text sized for a phone held at arm's length is enormous, fast cuts are more fatiguing, and the swipe is a remote control rather than a thumb. If a share of your audience is on a TV, opening frames built purely for phones are being tested in conditions they were not designed for.

Shorts grew a spine. On 23 September 2026 YouTube introduced Shorts series, which lets creators organise Shorts into seasons and episodes with custom thumbnails and sequential playback. For a channel whose Shorts are genuinely episodic, this converts a stream of isolated tests into something a viewer can continue, which is a different and more forgiving distribution problem. The full Made on YouTube round-up has the rest.

Part of the audience can now be switched off. On 14 January 2026 YouTube gave parents of supervised teen accounts the ability to cap daily Shorts time — from zero minutes up to two hours — or disable the Shorts feed entirely. For most channels this is noise. For channels whose audience skews young, it is a real reduction in available feed time, and it arrived without any change on your side.

The explanations that waste a month

Four popular diagnoses are worth ruling out early, because each one costs weeks and none of them is a mechanism.

Posting time. The Shorts feed is not chronological and does not clear a queue. A Short surfaces when a viewer's session reaches it, which can be days later. Publishing at midnight is not why the Short is flat.

Hashtags and keyword stuffing. Shorts are recommended, not searched for, in the vast majority of cases. A wall of hashtags in the description is not a lever; it is a habit carried over from a platform where it was never a lever either.

Deleting the underperformers. Removing a Short removes whatever information it contributed and none of the supposed penalty, because the penalty does not exist. Shorts are cheap enough to produce that the back catalogue costs you nothing to keep, and a Short that found 200 of the right people is worth more than a tidy grid.

Buying views to "trigger" the algorithm. Artificial traffic is explicitly listed among the views YouTube considers ineligible, and the audience it fabricates swipes away immediately, which is the one signal you most need to keep clean. The full version of that argument is a separate post, and the conclusion has not changed.

The diagnostic, in order

1. Discard the last two days of data. It is incomplete, and Shorts data moves fast enough to mislead you inside that window.
2. Confirm the upload is classified as a Short: square or taller, three minutes or under, Shorts feed present in traffic sources.
3. Check the audience setting at channel level, then at video level. Then check age restriction and country restrictions.
4. Record views and engaged views side by side for the last ten Shorts. Decide which failure you have before going further.
5. If engaged views are a small fraction of views, go to the first frame and the first second — this is a packaging problem.
6. If engaged views are nearly all of a small number of views, look at shown-in-feed. Low shown-in-feed with good swipe-through is an eligibility or coherence problem, not a content one.
7. Rank the last ten by swipe-away and by average percentage viewed. Look at the bottom three of each list and write down what they share.
8. Cut the next five Shorts 30% shorter than your instinct, with the promise in the first sentence and no outro.
9. Change one variable at a time, and give each change five uploads before judging it. Shorts are noisy enough that three is not a sample.

When the Short is fine and the channel is the problem

Sometimes every gate passes and the numbers still say no. At that point the honest reading is that the Shorts are competent and unremarkable — well made, well cut, and about something a stranger has no reason to stop for. This is the least welcome diagnosis and the most common one on channels that have been publishing consistently for months without movement.

It is also the only diagnosis that responds to effort in a predictable way. Premise beats execution in short-form by a wide margin, because execution is what stops someone leaving and premise is what makes them stay in the first place. If you are producing five Shorts a week from one idea a week, the constraint is upstream of every setting in this article: the ideas are being rationed and the edits are being over-served.

And it is worth being clear-eyed about what a fixed Shorts problem buys. Reach in the feed does not automatically become an audience — our look at whether Shorts actually help a channel goes through what the crossing to long-form does and does not deliver. Fix the Shorts so they get seen, then decide separately what you want the views for.

What to change on the next five Shorts

Run the gates once, properly, and you will usually find the answer is not where you assumed. The Shorts dying at zero are almost always classification or a setting. The Shorts dying at a few hundred are almost always the first frame. The Shorts dying at a consistent ceiling are almost always the premise. Those three problems share nothing, and treating all of them as "the algorithm" is why the fixes so often fail.

The one that transfers furthest is the packaging discipline, because it is the same discipline that governs long-form. A frame that reads instantly at a glance, one subject, a handful of words, and enough contrast to survive a small screen in bad light — that is a thumbnail brief, whether it is being used as a Shorts cover, a first frame, or a 1280×720 image on a watch page. If making those frames is the step where your process breaks down, Thumblore exists to make that part take minutes rather than an evening, on Shorts covers and long-form thumbnails alike.

From there, the work is unglamorous and finite: shorter cuts, sharper premises, one variable at a time, five uploads before a verdict. If you want the long-form counterpart to this piece, the views drop diagnostic runs the same discipline across impressions and click-through rate, and the Shorts cover guide handles the frame itself.