CTR without impressions is basically half a sentence
This is probably the biggest mistake I see when people talk about CTR. They give me the percentage. That’s it. “CTR is 9%.” Okay. On how many impressions?
Because 12% CTR on 4,000 impressions and 4% CTR on 400,000 impressions are two completely different situations. If we’re just grading thumbnails, 12% looks like the obvious winner. But the second video got shown to one hundred times more people. That changes everything.
The first package converted a small group really well. Great. But it still leaves me with another question: why did the room stay that small? And the thumbnail cannot answer that for me.
I stopped treating CTR like a thumbnail score
This was a big shift for me. I used to look at CTR much more directly. Good CTR? Good packaging. Bad CTR? Packaging issue. The more launches I’ve watched, the less useful that way of thinking has become.
CTR is not really telling me is this thumbnail good? It is telling me: of the people YouTube chose to show this video to, how many clicked? Those are not the same question. Because who YouTube chose matters. A subscriber is not the same as some random person scrolling Home. A returning viewer is not the same as somebody finding you through Suggested for the first time. Someone searching for your exact topic is already halfway sold.
So if a video is sitting at 10% CTR but most of that activity is coming from people who already know the channel, I am not immediately celebrating like we cracked YouTube. Maybe we did. Maybe we just showed the right video to the easiest possible audience. I need more information.
The first place I go is traffic source
I don’t touch the thumbnail yet. I want to know where the views came from. Browse. Suggested. Search. Notifications. Channel pages. External. Because a percentage without knowing where those impressions came from can be wildly misleading.
Someone who typed the exact topic into YouTube Search already has intent. They are looking for something. If my video appears and matches what they wanted, of course the click behavior can be strong. Now compare that with Home. That person was not looking for me. They were just scrolling. Now I’m competing with everything else YouTube thinks they might watch. That is a much harder click.
So if somebody tells me the video has 9% CTR, my next question is: where? Because 9% from a warm audience and 9% while YouTube is blasting the video across Browse are two very different accomplishments.
Then I want to see whether YouTube ever opened the room up
This is where it gets interesting. A video can start with a small group that is almost perfectly matched to it. Those people click. CTR looks fantastic. Then one of two things happens. YouTube starts expanding the audience, or it doesn’t.
If it does expand, I expect CTR to move. It might fall. That doesn’t automatically scare me. If we went from 10,000 impressions at 10% to 100,000 impressions at 5%, I am not looking at that and saying oh shit, CTR collapsed. I’m looking at the fact that YouTube just gave the video another 90,000 chances to get watched. The test changed. We are not talking to the same room anymore.
That is why I hate looking at one percentage without looking at what happened around it. The number can get “worse” while the video is actually doing something much more valuable.
And sometimes the idea is just small
This is another thing I had to learn. Not every video deserves a giant audience. That sounds obvious until you are staring at a video with a great CTR and wondering why YouTube will not move it.
Sometimes the answer is that the people who care about this really care about it. There just are not enough of them. That is not necessarily a packaging failure. It might be an idea problem. Or, more accurately, an idea-size problem. The thumbnail can be great. The title can be great. The video can be good. And the subject itself can still have a ceiling.
That is where strategy starts mattering more than another Photoshop pass. Because if the room is inherently small, changing the door does not make the building bigger.
This is why generic CTR benchmarks drive me crazy
I don’t really care that some article says 4% is good. Or 6%. Or 10%. Good for who? A 5% CTR might be incredible on a video YouTube is pushing aggressively into a huge cold audience. A 10% CTR might be completely normal for a channel’s subscribers in the first few hours. The number needs context.
And the best context I have is usually the channel itself. What does this channel normally do? What happens at this level of impressions? What happens when Browse starts scaling? What does Search usually look like? How do returning viewers behave? What does this channel’s audience consider clickable?
That matters to me way more than somebody else’s screenshot of a 14% CTR on X. Different channel. Different audience. Different topic. Different distribution. Why would I use their number to grade mine?
Then I look at what happened after the click
This is where I connect packaging back to retention. Because high CTR by itself can fool you.
Let’s say a video has high CTR, low impressions, and strong retention. Now I’m interested. The people who were shown the video clicked, and then they stayed. That tells me the video is working for the audience that received it. It still doesn’t prove YouTube is going to scale it. But now I’m much less interested in screwing around with the thumbnail. I want to know why the audience stayed small.
Now flip it. High CTR, low impressions, and people click and immediately leave. That is a completely different read. Now maybe the package is actually too effective at selling something the video does not deliver. Maybe we got the click and lost the person. That is not a win.
Again, this is why I keep coming back to the same thing: one metric cannot diagnose the whole video. It just can’t.
What I actually do when I see this
When I see high CTR and disappointing views, this is basically the order I work through it.
How many impressions are we actually talking about? First, I need to know whether we truly have a distribution problem. If YouTube barely showed the thing to anyone, I’m not going to pretend the CTR tells me everything.
Where did the impressions come from? Was this mostly subscribers? Search? Browse? Suggested? That changes what the CTR means.
Did the audience expand? Did YouTube start testing the video outside the obvious audience? And if it did, what happened? That is usually where the more interesting story starts.
What does this channel normally do? Not the internet. Not some guru. This channel. I want its own history.
Did people stay after clicking? If people click and stay, the package and video are at least aligned for that audience. If they click and leave, now I have another problem to solve.
Was the idea ever big enough? This one is uncomfortable because there is no thumbnail trick for it. Sometimes we chose a subject with limited reach. That doesn’t make the video bad. But it absolutely matters if the goal was growth.