In this post i'm going to show you how to beat the youtube algorithm in 2022. I'm going to show you draw dropping graphs. I'm going to show you data that's never been seen before. I'm going to reveal the algorithm secrets as told by youtube themselves.
And if you read this post to its conclusion. I guarantee you it will fundamentally change the way you think about the youtube algorithm. It's pretty exciting isn't it say hello to todd he works at youtube. And is an expert in search and discovery.
If you've ever watched the creator insider channel. Which you really should he often pops up on there to remind us that the algorithm. Isn't against you spoiler alert it is. And recently took his wealth of knowledge and data into the conferencing circuit.
Here he is talking at vidsummit. And he had some fascinating insights to share. But just before we jump into this. I'm curious to know if you could ask today one question about the youtube algorithm.
Beat the YouTube Algorithm
You're interested in right now that graph. I teased a little earlier on watch time. Let's take a look what is good audience retention on youtube videos. Well here we have it for a three minute video the average view duration is 62 for a 30 minute video.
It's half fat at just 31. Now i thought that youtube were trying to avoid these types of global bench marking. Because ultimately it just gives creators a stick to prod youtube. With when their expectations are not met.
Let's say for example that. I had a 50 minute video with a 45 average view duration. That doesn't take off that would be jolly annoying wouldn't it. Or on the flip side you have a video with very mediocre audience retention.
That would put it around about here on the graph. And yet it blows up performing well beyond typical of course. That isn't annoying in the slightest.
You want it to happen but at the same time. You're left scratching your head figuring out. How it happened.
Audience retention
This might help explain it a little bit watch time. For the video was above typical right out of the gate so youtube tested the video with new viewers. Who aren't as interested in your content. And that lowers the average view duration.
Relative audience retention is pretty strong too. But this is just one example up against the millions of videos that make up this data. So if you do want to match up your audience retention. With the rest of youtube. Now's your chance although having looked at all of this there is one big data point.
That's missing from this graph youtube shorts. Now full disclaimer this is not official youtube data but this is where i think youtube shops will probably end up. On that graph perfect should probably give this back.
Take it go on it's yours clearly. I haven't figured out the dog algorithm. Ever when it comes to youtube. It's pretty much impossible to take one single number from the analytics. And extrapolate success or failure and watch time is no different.
Watch Time
And that's because according to the youtube algorithm not all watch time is considered equal sometimes. A viewer can watch 30 minutes of content on youtube. And come away from it not feeling very satisfied.
And if that is the case it could lead to that person not using youtube for a couple of days. And opting for tick tock or instagram reels. Instead and that is what youtube hates they're in it for the long game youtube's algorithm.
Wants viewers to be satisfied. By what they watch today tomorrow next week next month next year next decade. So watch time simply isn't enough by itself youtube needs. To measure the value and satisfaction of the watch time.
And they do that with many other signals. If you want to dive into this more we've got a whole post over here. But what specifically mentioned in his conference talk. Was the idea of transformation experiences youtube can do this.
The YouTube Algorithm
Because it measures 80 billion different signals whatever that means. And when a video successfully transforms a viewing experience in a positive way. The algorithm wants to amplify that so yeah. Let's talk about that a little bit does youtube give your video a fair chance. To be discovered well according to today.
At youtube over 600 million different videos are recommended every day. On the home page and through suggested video. Now that sounds like a really incredibly impressive number. On the face of it but is it really.
I genuinely don't know for me this statistic lacks. A little bit of context since there are over 2 billion monthly users on youtube. I don't know if this includes all languages. There are millions of new videos published every single day.
And yet i can still be recommended one from 10 years ago. So 600 million so let's try and add a little bit of context to that. And this is something that todd said about viewers. The history of what a user has watched.
YouTube History
Is the most valuable thing that the youtube algorithm can understand. That's because a viewers recommendations. What they see on the home page up next suggested videos. Are highly influenced by personalization.
What youtube knows about the viewer so let me give you my personal experience of this. I don't know about you but. I have multiple interests and when i'm on youtube. I tend to associate different channels with each one of those interests.
I religiously watch for tech stuff. I watch your channel called wrestling bios for 1990s. Wrestling stuff and i watch oliver harper for anything to do. With film retrospectives and reviews now because. I watch these channels so regularly at least once a week.
These channels frequently appear on my home page. Even though i'm not subscribed to two of them. I gave up subscribing to channels years ago. YouTube has all of the signals it needs doesn't need this one. Anymore hasn't needed it.
For years now if one of those channels covered all of those topics. Would i just watch that one channel. I don't think so their styles personalities. And formats match their topic. So as usual in some respects. This goes back to channel focus and becoming the best creator for a certain niche.
And target audience on youtube. I have a deep passion for youtube tech wrestling films. But what do you come here to watch none of the other stuff. And when you come to the vid iq channel.
And really enjoy a video about youtube education. What do you really want to watch next from us. And that leads us on to another really insightful data point.
Thumbnail(beat the YouTube Algorithm)
From if 100 people watch your video this week 700 potential viewers will see your thumbnails. The following week or to put it another way. For every view you earn youtube. Gives you on average seven more opportunities to gain.
A another view this is basically youtube's advertising budget for your content. And with some back of a napkin maths. We can take the views from one week and the impressions from the following week. To get a ratio and it looks like youtube is even more generous.
Than average with its content advertising budget for the vid iq channel. However just like all of the other numbers. There's a really critical human element behind the data. Youtube isn't giving you seven random opportunities.
It's been really selective with which viewers they put your thumbnails. In front of i'm sure you've all seen this shelf before. When shopping on amazon what else customers buy after viewing an item. Amazon knows that if you intend to buy a spiderman mask.
You probably want to buy other marvel collectibles as well. I mean hello i have a thousand dollars worth of amazon. Influence decisions that's sitting on my shelves. And well youtube is pretty much the same but just a little more subtle about it.
Everything on this youtube homepage is driven by personalization not only from the creators. I watch on a regular basis but by the content youtube. knows other viewers tend to watch even. If i myself have never seen them before every time.
Somebody watches one of your videos and then they watch a another one of your videos. Or a suggested video by youtube. After watching one of your videos that strengthens the relationship between similar videos. Which makes session watch time average views per viewers.
YouTube studio
And this panel on the youtube studio which shows you. What your viewers are watching elsewhere on youtube. All the more critical you might call this passive collaboration. With other creators but at the same time you end up.
Competing with them which means that your titles and thumbnails. Are super important which leads on to click-through rate unfortunately. Wasn't able to provide any clutch information about click-through rate.
But there's good reason for that it's a constantly evolving fluid piece of data. Very hard to measure very hard to calculate as i demonstrated in this post. In the past youtube's algorithm has only been explained.
Through insane diagrams such as this so. I want to thank todd in his efforts. To make this type of stuff more approachable for creators. But how does all of this fundamentally change. The way you think about the youtube algorithm.
Well here's the simple answer the algorithm is less to do with numbers. And more to do with people in todd's own words. What the algorithm is designed to do is follow the audience. And make them happy and that is what todd youtube.
Conclusion
And myself recommend you do replace the word algorithm with audience. You came into this post wanting to know how to beat the youtube algorithm. Well i hope you're leaving this post wanting to know. How to beat the youtube audio no that doesn't sound right does it. How to build a youtube audience. And you can start doing that by removing these three rookie youtube mistakes. That everybody makes by reading this post over here.