Guide
How to get more views on YouTube
Every other page on this subject hands you fifteen tips. That is the wrong shape for the problem, because the fix for “nobody is being shown this video” is the opposite of the fix for “everyone clicks and then leaves”. Apply the wrong one and you make it worse.
Low views is not one problem. It is five, and three numbers you already have tell you which one you have. Get the diagnosis right and the fix is usually obvious.
First, the diagnosis
Open your most recent video that has had at least a week to settle. Read these three numbers in this order and stop at the first one that looks wrong.
- 1
Impressions
Is YouTube showing it to anyone at all?
Studio → Content → click the video → Reach
- 2
Click-through rate
Of the people shown it, are any choosing it?
Same Reach tab, next to impressions
- 3
Average view duration
Of the people who clicked, are any staying?
Studio → Content → the video → Engagement
The order matters. Click-through rate on a video with two hundred impressions is not a measurement, it is noise, so there is no point looking at it until impressions are real. The same goes for retention on a video nobody clicked.
The five problems
Problem 1
YouTube is barely showing the video to anyone
You see: Under a few thousand impressions, and the number is not climbing
Impressions are YouTube's decision, not yours. The system shows a new video to a small test audience, usually your subscribers and people with adjacent watch histories, then expands or stops based on what those people do. Low impressions that stay low means the test happened and did not go well, or that there was no audience to test it on in the first place.
For a small channel the second reason is far more common than creators assume. If the topic has no existing search demand and no adjacent audience already watching similar videos, there is nowhere for YouTube to put it. The video is not being suppressed. It has no natural home.
Check the Traffic sources tab before changing anything
If Browse and Suggested are near zero but Search has a trickle, the topic works and the packaging is the problem. If all three are near zero, the topic itself has no audience and no thumbnail will fix that.
Make the next video answer a question people already type
Search is the one door that opens for a channel with no track record. Browse and Suggested need YouTube to already trust you. A specific question, phrased the way a viewer would phrase it, can rank with two hundred subscribers.
Get narrower, not broader
Broad topics put you against every large channel. A version-specific or problem-specific title has almost no competition and a viewer with real intent behind it.
Look at whether anything similar exists and performs
If nobody has made a successful video on this, that is occasionally an opportunity and usually a signal that nobody wants it. Assume the second until you find evidence for the first.
The trap: Reuploading, changing tags, or rewriting the description. None of these affect impressions in any meaningful way. Tags in particular do almost nothing; YouTube reads your title, thumbnail, and the video itself.
Problem 2
People are seeing it and choosing something else
You see: Impressions are healthy, click-through rate is low relative to your own average
Click-through rate is a comparison, not an absolute. Your video appeared next to five other options and lost. That is the whole mechanism. It does not mean the thumbnail is ugly; it means the offer was weaker than the alternatives on screen at that moment.
This is also the number most distorted by advice. YouTube's own guidance is that there is no good universal CTR target, because it varies enormously by where the impression happened. A video shown mostly on your channel page will show a very different rate from one shown in Suggested next to a large channel.
Compare against your own channel, not against a number from a blog
Your median CTR is the only meaningful baseline. A video at 4% on a channel that averages 3% is doing well, whatever a listicle told you the target was.
Say a different thing, do not restyle the same thing
Most thumbnail iteration changes colours and font weight, which changes nothing. The offer is what moves the number. A new angle on the same video beats a cleaner version of the old angle.
Check it at the size it is actually seen
You are designing at full size and it is being judged at roughly the width of a thumbnail on a phone. Text that reads perfectly in the editor becomes a grey smear. Two or three words is usually the ceiling.
Make the title and thumbnail do different jobs
When both say the same thing you have spent two slots on one idea. The thumbnail should raise a question the title answers, or the reverse.
Test on videos that already have impressions
Changing the packaging on a video nobody is being shown teaches you nothing. Iterate where there is traffic to learn from.
The trap: Chasing CTR at the cost of accuracy. A thumbnail that overstates gets the click and then loses the viewer in the first thirty seconds, which teaches YouTube that people who click are not satisfied. That is worse than a lower click rate.
Problem 3
The video is not delivering what the packaging promised
You see: Click-through is fine, average view duration falls off a cliff early
A steep early drop is almost always a mismatch between what the viewer expected and what the first thirty seconds gave them. They clicked for a specific thing and got an intro, a channel trailer, a request to subscribe, or a slow lead-up to the thing they came for.
The second common cause is that the promise was answered immediately and completely. If the title asks a question and the first line answers it fully, there is no reason to keep watching. The answer needs to arrive, but it needs to open something.
Open on the thing they clicked for
No introduction, no channel branding, no housekeeping. Deliver the first piece of the promise inside the first sentence and earn the rest of the runtime.
Watch your own retention graph and find the exact second
The graph tells you where people leave, which is more useful than any general advice about pacing. Watch the ten seconds before each drop and you will usually see the reason immediately.
Move the ask to the end
Asking for a subscription before you have delivered anything is asking a stranger for a favour. It reliably costs retention at the point where retention matters most.
Re-raise the question every minute or so
Retention is a series of small decisions to keep watching. Each one needs a reason. A video that answers everything by minute two has nothing left to hold anyone.
The trap: Adding jump cuts and sound effects to fix pacing. Editing speed does not fix a structural problem. If the viewer no longer wants the answer, cutting the pauses just delivers the same disappointment faster.
Problem 4
Nothing connects one video to the next
You see: Individual videos do fine but the channel is not growing
YouTube grows channels by learning who to show them to. That requires consistency in subject, format or both. If every video serves a different audience, each one starts from scratch and none of them build on the last.
The other version of this is a channel with no reason to return. Videos that are complete in themselves and unconnected to anything else give a satisfied viewer no next step, so they leave and the relationship ends there.
Pick a viewer, not a topic
"Videos for people learning DaVinci Resolve on a laptop that struggles with it" is a channel. "Video editing" is a category. The first one tells YouTube exactly who to test each upload on.
Make the next video the obvious sequel to the last
Series compound because a satisfied viewer has somewhere to go. A single video can go viral; only a sequence builds an audience.
Repeat the formats that worked instead of moving on
Most creators abandon a working format after two videos because they are bored of it. Your audience is not bored of it; they have seen it twice.
Look at which videos brought subscribers, not which got views
Studio shows subscribers gained per video. A video with modest views that converted well is a better template than a high-view video that converted nobody.
The trap: Concluding you need to post more often. Publishing three unconnected videos a week compounds no faster than one, and usually less, because the quality drops and none of them tell YouTube anything consistent.
Problem 5
Not enough videos yet for anything to be measurable
You see: Everything looks reasonable, the numbers are just small
Under roughly twenty videos, most of what you are looking at is noise. One video landing well or badly moves every average you have, and there is not enough history for YouTube to have learned who your audience is.
This is the least satisfying diagnosis and the most common one. There is no adjustment to make, because the problem is not a mistake. It is a sample size.
Keep the format constant long enough to read the result
Changing everything after each upload means you never learn which change did what. Hold one variable, change one, and give it three or four videos.
Judge on batches of five, not on individual videos
Single-video variance on a small channel is enormous. Five videos of the same format is the smallest set that tells you anything.
Spend the time on picking topics, not on optimising
At this size the difference between a video that works and one that does not is almost always what it is about. Optimisation matters later, when there is traffic to optimise.
The trap: Treating a single flop as a signal and rebuilding the channel around it. Early data is mostly noise, and reacting to noise is how creators end up with twelve videos in nine formats and no audience.
Things that will not fix it
Each of these is taught constantly and each is either wrong or far less important than the attention it gets.
Posting at the right time is what unlocks views
YouTube's own position is that publish time is not known to affect a video's long-term performance. Publishing while your audience is awake can help the opening hours, and that matters for a live stream or a news reaction. For an evergreen tutorial it is close to irrelevant, and it is the single most over-taught tactic on YouTube.
Tags are how YouTube understands your video
YouTube reads your title, thumbnail, description and the content of the video itself. Tags exist mainly to catch common misspellings. Filling five hundred characters of tags is a ritual, not a strategy.
You need to post daily
Volume compounds only when the videos are consistent enough to teach YouTube who to show them to. Daily uploads of unrelated things compound slower than weekly uploads of the same thing done well, and burn you out on the way.
Keyword-stuffing the description helps you rank
The first hundred characters or so appear in search results and are worth writing carefully. A wall of keywords below the fold is read by nobody and does not help. If search traffic matters to you, the phrase belongs in the title and in a natural opening sentence.
Longer videos always earn more
Longer videos can carry more ad breaks, so the ceiling is higher. But a padded video loses viewers, and watch time lost is worth more than an extra mid-roll. Length should be decided by the subject, not by a monetisation target.
Subscribers are the number that matters
Most views on most channels now come from people who are not subscribed. Subscriber count is a lagging vanity metric; views per video and traffic sources tell you far more about whether anything is working.
The uncomfortable part
For most channels below a few thousand subscribers, the honest diagnosis is problem one or problem five: the topic had no audience, or there are not enough videos yet to read anything from. Neither is fixed by optimisation, and both are unsatisfying answers, which is why almost nobody writing about this gives them.
The single highest-leverage decision on YouTube is what the video is about, made before anything is filmed. Packaging can win a video that deserved to win. It cannot create demand that was never there.
Questions
Why did my views suddenly drop?
Usually one of three things: a video that was carrying your Browse traffic aged out, you changed format and the audience YouTube had learned to show you to no longer matches, or the drop is seasonal. Check whether the fall is across all videos or concentrated in one, because a channel-wide drop and a single video cooling off are completely different problems.
How many views is good for a small channel?
There is no useful universal number, because it depends entirely on how many people could plausibly want the video. A tutorial for a niche piece of software doing four hundred views may be reaching most of its available audience. Compare against your own median and against the topic, never against a benchmark from an article.
Do Shorts help or hurt long-form views?
They mostly do not transfer. Shorts viewers and long-form viewers behave differently and YouTube largely treats them as separate surfaces, so a Short going viral often adds subscribers who never watch a long video. Shorts are worth doing for their own sake, not as a funnel.
Should I delete videos that performed badly?
Almost never. A video with few views is not dragging your channel down; there is no channel-wide quality score being averaged. The exception is a video that actively misrepresents what your channel is now, which can confuse the audience-matching.
How long before a new video finds its audience?
Most of a video's lifetime views arrive over weeks and months, not in the first day. YouTube can keep surfacing something years later. Judging a video by its first twenty-four hours is the fastest way to draw the wrong conclusion.
Does changing the thumbnail after publishing reset anything?
No. Changing a thumbnail or title does not restart the video or penalise it, and on a video that already has impressions it is one of the few genuinely useful experiments available to you. On a video with no impressions it teaches you nothing.
Related