I opened a channel's analytics last month and the number that mattered was not views. It was 412,000 impressions on a video with 9,800 views. That is a 2.4% click through rate, and it means roughly 402,000 people saw the video, looked at it for about a second, and kept scrolling. The video was fine. The packaging lost.
That gap is the YouTube impressions funnel, and once you can see it as a funnel instead of a scoreboard, you stop guessing about growth. Every video moves through the same four narrow gates: impressions served, clicks earned, the first 30 seconds survived, and watch time banked. Each gate feeds the next. Fix the wrong one and nothing moves.
What is the YouTube impressions funnel?
The YouTube impressions funnel is the path a video takes from being shown on a screen to being watched long enough for YouTube to show it again. Four stages, in order.
- Impressions. YouTube shows your thumbnail somewhere: home feed, suggested column, search results, subscriptions tab. This is the only stage YouTube fully controls.
- Click through rate. Someone decides in under a second whether your thumbnail and title are worth their next ten minutes.
- The opening. The first 30 seconds decide whether the click was honest. This is where the promise in your packaging either lands or breaks.
- Sustained watch time. Average view duration and session behaviour tell YouTube whether to buy more impressions on your behalf.
The funnel is a loop, not a line. Good performance at stages two, three and four buys you more of stage one. That is why a video can sit at 3,000 impressions for a week and then suddenly hit 90,000 in a day. Nothing changed about the video. The system simply got enough evidence to widen the top.
Why do impressions stall when the video is good?
Impressions stall because YouTube stopped getting a reason to keep spending them. The algorithm is not judging quality in any human sense. It is running a very cheap experiment: show this thumbnail to a small slice of a likely audience, measure what happens, decide whether to expand.
If that slice clicks at 2%, the test ends quietly. If it clicks at 7% and people stay, the test expands. You do not get a notification either way.
This is the part most creators misread. They see flat impressions and conclude the algorithm buried them. Usually the algorithm offered a fair first test and the packaging did not convert it. YouTube's own creator resources are blunt about this: impressions are supply, and click through rate plus watch time is what decides whether supply keeps flowing.
Impressions are not a reward for making the video. They are a loan. Your packaging and your first minute decide whether YouTube lends you more.
How many impressions do you actually need?
You need fewer than you think, if the conversion is right. Run the arithmetic both ways.
A video with 400,000 impressions at 2.4% gets 9,600 views. A video with 120,000 impressions at 9% gets 10,800 views. The second video got less than a third of the exposure and still won. And because its click through rate is strong, YouTube keeps buying, so the second video probably ends the month at 40,000 views while the first one is done.
That is the whole argument for treating packaging as leverage rather than decoration. Going from 2.4% to 4.8% does not double your views. It doubles your clicks, then compounds, because doubled clicks earn more impressions, which earn more clicks. I wrote a longer breakdown of the benchmarks in what a good CTR actually looks like, including why the sitewide averages you read on Reddit are close to useless for any specific channel.
Two numbers worth pinning to your wall:
- Impressions with a bad CTR are a diagnosis, not a failure. 400,000 impressions means YouTube found an audience for your topic. That is the hard part solved.
- A video with 4,000 impressions and 12% CTR has a different problem. That is a distribution problem, usually topic or title framing, not a thumbnail problem.
Knowing which of those two you are looking at saves you months.
Where in the funnel is your problem, exactly?
Pull up any video and read three numbers together: impressions, click through rate, and average view duration. The combination tells you where to work.
Low impressions, high CTR. Your packaging is working on the people who see it. The topic is too narrow, or the title is missing the language people actually search and browse for. Work on subject selection and title framing. This is common on well made tutorial content, which is why packaging for tutorial videos has to fight on search intent and not just visual punch.
High impressions, low CTR. The classic. YouTube believes in the topic and nobody believes in the thumbnail. This is the cheapest fix on the list and the one most channels ignore for a year.
High CTR, low retention. Your packaging is overselling. The click was earned by a promise the first 60 seconds does not keep. Ironically this hurts you more than a weak thumbnail, because YouTube reads the bounce as a bad recommendation and pulls back impressions across the channel.
Everything low. Do not repackage yet. The topic itself is not wanted. Change the subject, not the graphics.
Most channels I look at sit squarely in the second bucket, and they have twenty or thirty videos sitting there at once.
Can repackaging old videos really reopen the funnel?
Yes, and it is the cheapest growth most channels have available. Your back catalogue is already indexed, already has watch time history, and is already eligible to be shown. It is just being shown with packaging that lost its argument.
When you replace a thumbnail on an older video, YouTube does something useful: it starts a small fresh test. Impressions that had flatlined get another slice. If the new thumbnail converts better, that slice widens. You did not shoot anything, write anything, or edit anything.
Here is how I prioritise a back catalogue, and the order matters:
- Highest impressions, lowest CTR. Maximum leverage. The audience is already being offered the video.
- Videos with strong retention but weak clicks. The content is proven. Only the door is closed.
- Evergreen topics that still get search traffic. Anything time bound is not worth the money.
- Videos that sit next to your best performer in suggested. They inherit impressions.
Eight videos is usually the right first batch, which is not an accident: it is exactly what our pricing and services page is built around. $250 for 8 custom thumbnails works out to about $31 per video, one thumbnail per video, unlimited revisions, a dedicated designer, and title consultation included. Turnaround runs 48 to 72 hours per thumbnail. If you want to test two directions on your biggest video, an A/B variant is $15, a title pack is $25, and rush is $20.
Spending $250 to reopen the funnel on eight videos that already have impressions is a very different bet from spending it on one new video that has none.
Why does the title matter as much as the thumbnail?
Because a viewer never sees them separately. The unit of decision on YouTube is the package: thumbnail, title, and the first frame of the video working as one argument. Judging a thumbnail alone is like judging a headline with the photo cropped out.
The practical rule I use is that the thumbnail and title should never say the same thing. If the thumbnail shows a wrecked kitchen, the title should not say "my kitchen got wrecked". The image supplies the situation and the title supplies the stake, or the reverse. Redundancy wastes the only second you get.
This is also where reaction and commentary formats live or die, because the whole promise is emotional rather than informational. We break the mechanics down in the guide to thumbnail design for reaction channels, and the same logic applies to podcast clips and audio led content, where podcast cover art design has to carry a promise with almost no visual event to work from.
That combined view is the reason I call this work packaging rather than thumbnail design. Thumbnail design is a deliverable. Packaging is a decision about what the video is claiming to be.
What does good packaging work actually look like?
It looks like an argument, then a file. Before anyone opens Photoshop, three questions get answered.
Who is this being shown to? A thumbnail for a home feed browser and a thumbnail for a search result are different objects. Search wants clarity and keyword match. Home feed wants curiosity and contrast.
What is the single readable idea? One idea. At 246 pixels wide on a phone, two ideas is zero ideas. If you have to squint at the mobile preview, it failed.
What is the promise, and can the video keep it? Overpromising is the most expensive mistake in the funnel because it damages the channel, not just the video.
Then the craft: contrast against YouTube's dark and light interfaces, faces with an actual legible expression rather than a generic open mouth, text kept to three or four words when text is used at all, and a colour that does not appear in the videos surrounding yours in the feed. Google's own YouTube Help documentation covers the technical specs, which are the easy part. The hard part is the judgement.
Every thumbnail we make is custom for that video. No templates, no swapping a face into last month's layout. Styles get tested on our own channels before they go near a client channel, which is also why we can tell you when a look is played out. You can see the range in our work, and if you want to see how a remote studio actually operates across timezones, the notes on running a YouTube packaging agency from Dubai explain the async workflow.
What should you do this week?
Open YouTube Studio, sort your videos by impressions, and write down the five with the highest impressions and the lowest click through rate. That list is your funnel leak, ranked by how much it is costing you.
Then decide honestly which of the four stages is actually broken. If it is stage one, no thumbnail will save you and you need better topics. If it is stage three, fix your openings before you spend a dollar on graphics. If it is stage two, and for most channels it is, you are looking at the cheapest fix in your entire operation.
We are Dubai based and work worldwide and async, so the timezone rarely matters. If you want a second opinion on which of your videos are leaking impressions, send me the list at samy@pixelsamy.com or through the contact form on the site, and I will tell you straight whether repackaging is worth it for your channel.