Previous: Am I OK?
Next: The Weirdest Place I've Ever Changed a Diaper



View count:36,442
Last sync:2024-06-05 23:45
People are talking about subscriptions and what they do and why they matter so I wanted to go in and look at why YouTube might be worried about subscriptions and what the big ol' deal is.

 (00:00) to (02:00)

Hello! We're gonna talk a little bit about, uh, the analytics of the vlogbrothers and Complexly networks, here, 'cause we have the opportunity to do that and we're talking a lot about subscriptions and sub boxes these days, and we.. I do wanna give some context to that. I'm not gonna talk about my opinions on the sub box changes, cause I made a video about that on vlogbrothers and you can go check it out, and probably you're coming from that video to watch this one.

So, here is the traffic sources for everybody on the vlogbrothers network, so we have a bunch of different channels that are all.. SciShow, CrashCourse, Healthcare Triage, Sexplanations, Animal Wonders, uh, new shows that you probably haven't heard of like Nature League, and how much the different things matter to the different kinds of channels. So, right here we're seeing where all of our impressions, our views, our watch time comes from, so we're actually looking at watch time right now, this is basically the exact same graph for views, uh, but that's a little bit less of an abstract number, so, uh, we see that this base level here is.. are my suggested videos, and we see that this red bit here is the browse features, and then we've got search, and channel pages, and other YouTube features and et. cetera. So, um, our biggest hunk, at 41% is suggested videos, that's prob.. usually about the same for a lot of YouTube channels, um, search is a big piece for CrashCourse specifically, and to some extent SciShow, browse features, though, is where, um, subscriptions is hiding. And it's really frustrating that subscriptions and home aren't on this as a separate category... I don't really know why, I guess it's so that you can see tha- th- the browse feature as being, sort of, like, what YouTube is doing for you as the category of, like, people actively on their, sort of, like, front-page-home experience of YouTube, whether that's subscriptions or home, but I find it frustrating because then we have to say, well, 38%, and then we click on browse features, and then we see what percent is home and what percent is subscriptions.

 (02:00) to (04:00)

So 65 percent is home so 65 percent of the 35 percent is coming from the home page and 27 percent of that. I forgot the I forgot the number already but 27 percent of that is uhhh coming from subscriptions, so that's like not a ton, but I mean look at it in terms of views it is a ton it's uhhhm 2 million views - almost 3 million views.

So it is a big deal uuuhhm and you can also see the click through rates in those places, the click through rates for home are higher than they are for subscriptions and this is probably what youtube is looking at and saying like home is doing a better job than subscriptions getting people what they want.

They're clicking - they more likely to click on a video we served them on home than on subscriptions because subscriptions is gonna be so much content and not really well divided, so that's that's one part of it. Another part of this, let's look at the vlogbrothers analytics for, oh god which one is it? It's not that one. I've got a bunch of other things up here. I put it right in the middle you guys that makes a lot of sense.

Let's look at the vlogbrothers breakdown between subscriptions and home, so these are the browse features over the last, so I don't know whats going on here uhm uuuh this this don't- ignore this green area that's a glitch, uuhm but we've got our so that that purple is subscriptions, and you'll see this is 2014 this is 2013 uh you'll see that this is seems to be on the decline, in fact it seems to be really on the decline lately and uuhhhm and I don't and it would be interesting.

Let's just go ahead and type in scishow here and see of if that holds up at all across a different channel, now this is gonna take forever to load so we're gonna cut now. I have made this display slightly differently for scishow. Most of this data is not useful down here because this is where we started to upload seven videos a week so there was a big jump in viewership unsurprisingly.

But what we see is from that time a you know like this is a very bumpy graph of how home is doing for us.

 (04:00) to (06:00)

Because that really depends on individual videos. But it's a pretty steady decline in terms of what, uh, subscriptions is doing for us. So I think that Youtube might be concerned about the fact that subscriptions seems to be, being a less useful product. And people not, not using it as much. Um, now I want to say, another thing about subscriptions here, uh, about why they are important beyond just, like, how- when and how people get the viewership. 

And I will say that it's been pretty steady on Scishow. So we've got, uh, 9 million views coming from subscriptions here. And we've got 8.5 million views coming from subscriptions here, so it's not that different over the last year or so. 

Let's check out Healthcare Triage and see how different that is. So again, Healthcare Triage also showing a really significant, actually, decline in views coming from subscriptions. Um, particularly as maybe a percentage of the views coming from home. But again, like these home views are very unstable, whereas subscription views are much more stable. 

It looks to me like Youtube is worried that the subscriptions product is just not being used as much by people. Unfortunately we don't have impression data. Uh, even if we did have impression data that might not be super useful because it wouldn't be like impression per like user, it would just be impression. So there might be more users on the platform, a few of them using subscriptions. 

But we do see a pretty, like, Healthcare Triage sees a really substantial decline. I mean like, I'm very interested in other people's data. In fact, if you want to share data on this stuff, it's pretty, uh, you can look at this and sorta figure out how I found this data. Um, and I'd love to see, uh, see that kind of information shared, whether that's at, you know, sort of like, internally between friends, or if you just want to go ahead and put a #youtuberanalytics on Twitter and show us what your subs impressions and numbers are looking like. 

 (06:00) to (08:00)

You may not be into, uh, the impressions beta, so you may not have that data. But hopefully you will be soon. But I also wanted to talk a little bit about why subscriptions are important and, uh, a little bit about how they function. 

So here are two vlogbrother videos over here. We've got a video that didn't do well, this video is about red eyed tree frogs. Um, I don't know, it got a hundred and something thousand views. And then we got my marijuana video that probably got close to 300,000 views. Uh, and we're often asking "what's the difference between those two kind of videos?"

Well this is a very clear signal. Uh, this is just the data for the day that the video came out. These video's got similar number of impressions. They're not exactly the same, a little bit more on the marijuana video, um but you'll see the impressions click through rate here is 11.2 percent on this first day. Whereas red-eyed tree frogs is 6.2 percent.

Now I'm not saying that's like a bad thing for the vlogbrothers channel. Frankly, I think it's good to have content that isn't designed to get a bunch of clicks every once in a while. Or even most of the time. That results in, like you know, you have roughly the same number of impressions, but you have far more, uh, viewership and minutes watched. 

 And if you go down here to where those views are happening on this marijuana video, 8 percent of the subscribers, and this number 555,000 is at the moment at least, and they say that- even with this new subscriptions thing that there will still be those impressions. These videos won't leave people's subscription feeds. That like this number, at the moment, is a steady number. Like that's just the people who opened subscriptions and my thumbnail is there. 

So unfortunately, I can't see how that number has changed over time. Uh, it may be that people are using subscriptions less, and that's something that's concerning Youtube. But there is this 555,000 number, uh, that is gonna be roughly the same. And if we look at this video, uh, 536. It might be that John uploaded a little later in the day, it could be that Tuesday just had fewer views or whatever that is. 

 (08:00) to (10:00)

But then you see our click through rate. So the click through rate for this video on the subs page is 4.5 percent and on the homepage is 12.15 percent, which is actually really high. But this video did better. 8 percent, almost 9 percent, click through on subscriptions and 17 percent click through on the homepage. 

Now that's signaling to Youtube that it should be sharing it more, and as time goes on it does indeed share it more. And I have an example of the same phenomenon from, uh, Scishow here too. How Computers Find Naked People in Photos, this is the kind of thumbnail that people are gonna see, and uh, and this other video that did very poorly, one of our worst videos in a long long time, Your Asthma and Allergies Aren't Causing Mental Illness. Which has both a- mostly a title problem I think.

I don't think that people thought they were. Um, and so that was a misfire on our part. But yeah, our click through rate is in the single digits like 1/2 percent on browse features, and I did not open the subs verses home tab but yeah, 1 and 1 percent here. And then over here on How Computers Find Naked People, one, we're getting very good click through rate on suggested video which is why this video has continued to grow because it's doing well on suggestions and in side bars.

And then if you go to browse features, you'll see that it also did quite well with subscribers. Um, now it's important to note that this click through rate is much lower, and always Scishow's click through rate is much lower, than Vlogbrothers. And so you have to sort of ask yourself, like if Vlogbrothers, every single Vlogbrother's video has a higher click through rate for both home and subscription. Why do Scishow videos so often do much better than Vlogbrothers videos? 

And my interpretation of this is that videos aren't necessarily compared across channels. I think they are to some extent of course, but I think they are also compared internally. And so you look at a Scishow video- like Youtube's algorithm is looking at all Scishow videos, and it's saying "well this video got better clicks and better watch through." Six minute of watch through is very good. 

 (10:00) to (12:00)

Because this is a longer video than the average SciShow video, not than the average video. Um, I think that, like, there are lots of reasons why SciShow click-through rates are lower. I think, one, it's because they show up more, there are a lot more of them. I think it's also because people who are subscribed to SciShow tend to be subscribed to a lot of channels and sort of have less, like, deep brand affiliation with SciShow. 

You also see a lot more impressions here. Uh, so, oh shoot, well, none of this data matters at all you guys. You gotta look at just the one day. You've gotta look at just the 2nd of May, and then this will be actual data that we can compare. So maybe I was all wrong. Uh, but no, I wasn't.

So obviously, if we just look at the day the video came out, those numbers are a lot higher, because they're people who, like you know, are regular viewers. But we can see much higher impressions because SciShow, there are more SciShow subscibers and also they're more active because they're newer subscribers than vlogbrothers subscribers, for you know, twelve years now.

Um, and you, but the home, home is interesting because that's not subscribers; that is YouTube deciding to put that video there. And then if we go to our video that didn't do as well, our allergies video, you'll probably see that it didn't end up on home at all because people weren't clicking on it when it went there. So we have to go to April 27th. Let's see what the impression rate jumps up to here, if we can get into 2 or 3% click-through rates. Nope. We can't. Um, so this is a lesson learnt. See, 800 thousand impressions, just like anything else, but then that's not going to push a lot to home. YouTube's getting some data here, and finally people aren't clicking on it in the way that they... even they are. They're pushing it to people who are more likely to click on it than in their subscriptions, and so all of this is data that is indicating to YouTube that subscriptions is a less and less useful product. But, I also think that subscriptions is kind of supposed to be one thing specifically.

 (12:00) to (14:00)

And it is a little bit of a contract both with creators and with viewers that they are signing up to see content and they are signing up to see the content in the way that they expect to see it. That maybe means that YouTube isn't solving a problem that they could otherwise solve but sometimes a problem exists and you don't solve it because it is a feature that a lot of people sort of have a cultural expectation of. 

And also like my subscribers, like these people, maybe not the people who clicked on that particular video, but the people who click on the, these folks who saw these 29,000 people who saw this red-eyed tree frog video. And even though it's not a very good title for a video they clicked on it because they're our super-duper core audience that is going to look for our content in the place where we have told them our content is. Now, I understand YouTube's concern about the sort of wane of subscriptions and using a lot of valuable real-estate on and also on the App to push to subscriptions, but I think that subscriptions is a very important product. And I think there are ways to refine that product and to make it into a more useful experience that still do not make that an algorithmic experience, that still do not take videos and serve them to people out of order. And I talked about that some in my vlogbrothers video. So thank you for listening and watching this dive into YouTube analytics. I know that you love it, like all these numbers. And thanks so much for making vlogbrothers clickthrough rates so high and caring a ton about our content and being just great people in this community and we really appreciate it and are super honoured by that. We know that it is unusual

 (14:00) to (14:06)

And it matters a lot, so thank you. And that's all. Bye bye.