 One thing that we always talk about, I think that goes unnoticed though, is for as much as everyone here at VidCon talks about the algorithm, one thing that tends to go overlooked is actually the psychology of the viewer. Because at the end of the day, yes, the algorithm ultimately dictates what you do and don't see in your video feeds, in your subscription feeds, and things like that. But at the end of the day, this is a human-driven platform as well. The viewer has to click on that video. The viewer has to want to engage in that piece of content, and they want to stick around and watch with you and engage with this sort of content that you're producing on this platform, right? And so today, that's what I wanted to dig down into, because human psychology plays a huge role in every step of this process, from the discovery of your video, to the thumbnail, to the way they process the title, down to the content when they're actually watching it, and finally to the end screen call to action, where you're asking them to do something for you. And if you understand and have a solid understanding about how the human brain operates and what it's thinking and what processes are going on in each step of that journey, you're more empowered to create content that's going to be better optimized for not just the algorithm, but how humans behave with that content. So first, let's start off with the actual discovery, and I want to find a piece of content that's going to attract me online, right? So the first principle that really guides expectations on this platform is known as priming. It's this idea that we as humans are primed to expect certain things when we're thinking about a certain piece of content or a certain genre of video. A great example of this is in-movie theaters right now. There's this movie called Hereditary, which is this crazy horror movie that plays with the primed expectations of humans about what happens in a horror movie. And midway through the movie, it subverts that expectation, and you're taken in a completely different direction. YouTube videos function the same way. If I'm looking for a tweenage challenge-based content, then I'm expecting certain tropes in those thumbnails that are immediately queuing me in, even if I don't realize that this is a tweenage challenge type video, and that's everything from the shapes that I'm seeing on screen, the colors that I'm seeing in that thumbnail, the faces that are used, the expressions, things like that. And so here I bring up two examples. We have tween-based challenge content. We've done analytical analysis of this across the platform, and the colors light, sky blue, pastel colors actually function and trigger the brain, the viewer, to think, oh, this is tweenage challenge-based content. This is the sort of content that really fits in alignment with that content category. Putting a white-stroke border around the personality, the character, the key image on that frame, that's going to help you in the audience that this is what's important there. Minecraft channels or other video game channels, again, we've run analytical studies on this, and we found that using official key art or official graphics from the gameplay themselves, things that have been released by the video game companies actually perform better than if you're using fan art or stuff that is very similar to official key art, but it's very clearly not an official asset. And so these are the sorts of things that are immediately queuing in, the viewer, oh, this is that style of video, this is what I expect from this content vertical, I feel comfortable clicking on this video because I know what is on the other side is something that I expect, because that's another guiding principle of humans is that we don't like to click black boxes. We don't like to click mystery boxes online. If we're clicking on a piece of content, we want to know what we're getting, and these are the sorts of cues immediately when I first see your video thumbnail in the suggested video feed are immediately queuing me in. So I pulled out this thumbnail from Good Mythical Morning, Rhett and Link. It answers a lot of the immediate questions. It's telling you as the front door of this video what you can expect to see here. You know, you have the clear branding in the bottom left-hand corner. You have Rhett and Link's faces. You can tell that this is fun kind of off-the-wall wacky comedy. This is immediately just from seeing this image. I know what to expect from this video. I am primed using the bright colors, the wacky positions, the branding in the corner that I know what I'm going to roughly get when I click on this video. The other reason why I picked this thumbnail though is it illustrates a couple of other interesting features. So first, you'll notice that it has a white background. Think about your suggested video feed. Think about how you see here at YouTube homepage. Everything is a box. Everything is a rectangle. If you're in the gaming sphere, a lot of them are blackish or dark rectangles. Having a white background actually allows you to blend into the background of YouTube. And so now that this thumbnail and any thumbnail that has a white background no longer is just a rectangle. And so as I'm scanning through the hundreds of thumbnails that are presented to me on a day-to-day basis, now all of a sudden that one jumps out to me because it's a different shape than all the others. Again, my human brain, my mind is triggered by the idea that, oh, this is different. My eye is drawn to the thing that's different and I focus in on it. And that makes me that much more likely to click on that video because it caught my attention in a sea of thousands of other things that are competing for my attention. The one thing I will ding them on that we'll talk about here in a second is that their faces are upside down and humans actually have a difficult time processing upside-down faces. You do want to include faces in your thumbnails, but upside-down faces actually don't have the same boost that a right-side-up face actually features, which we'll talk about. So we've done testing on this, and I can say for certainty that people are, when they consume a thumbnail, they consume it, at least in English-speaking territories, that they consume the thumbnail like they read a book. They're primed to absorb the image in the same way that you would kind of read a piece of text, right, where they're starting in the upper left-hand corner and processing it over to the right, or they're starting in the upper left-hand corner and processing it down and diagonally to the right. So here, this is one of my favorite examples from our own channels, where very clearly there's a story being driven where Mario is looking into the thumbnail, or he's the first thing that you see, and then his gaze and his body position are driving you down to follow the arm, and then the words guilty are kind of adding to that. It's important to note that the left side of the screen in thumbnails is stronger and more important than the right side of the screen. Again, because of this idea of people start consuming content on the left and proceed to the right, think about how you consume the watch page on YouTube. I'm watching my video here, and I move my eye gaze over to the right a little bit, and that's where I'm finally seeing those thumbnails. So the first impression of your video that I'm getting is on the left side of it. So you want whatever your key image, the most important thing, the most impressive thing, the thing that you expect to really get the audience excited about your video should be on the left-hand side. The other thing to note here about this kind of how humans consume thumbnails is a lot of brands, a lot of creators will use branded logos. You saw it in the Good Mythical Morning thumbnail with the GMM in the bottom left-hand corner. Where do you put that branding in your thumbnail? Well, again, think about this. If the first thing that I'm seeing is in the upper left-hand corner of the screen, you don't want it to be necessarily your brand because my eye sees that and it stops. We've run tests on this where if we put the logo in the upper left-hand corner of the screen, the eye stops at that corner and doesn't actually, because it serves as a wall almost, and so they're not going down deeper into the thumbnail to consume the rest of the image. If you're putting your branding logo in any corner, you want it in the upper right-hand corner because usually that's one of the less important corners of a thumbnail, or the bottom, upper right or bottom left. You don't want it in the bottom right because that's where your timestamp is, and so your branding is going to get covered up by the timestamps on YouTube. That's a little bit about that. Ideally, and so we'll actually talk about that in the slide here, but humans are attracted to faces. I already mentioned this, but if you go across YouTube's platform, if you look at the data across the board, human psychology is just, we empathize with faces. We are drawn to our own face. We are drawn to the faces of others, and so thumbnails that have faces that are large and expressive and interesting are ultimately more clickable and more attractive as, again, I'm scrolling through my video feeds and those that don't have those sorts of emotions or faces. And again, you see this in the data where each step of this you kind of see more and more click through. So a basic face, excuse me, a basic face is better than a thumbnail with no face. A face with emotion is better than a basic face. A face with emotion that's driving audience into the thumbnail is the best of all. So this is actually one of my favorite thumbnails on YouTube in my years of doing YouTube because it does so many things right here. So you have the person on the left, right, left. Their eyes are actually driving the consumer, the viewer's gaze in deeper into the thumbnail so that way I want to consume the entirety of this image. You have a clear emotion on both faces. You have a kind of surprising or disgusting image here that is just inherently attractive on the platform. It's an interesting thumbnail. It's a half-completed action. And the eyes actually here, those aren't her complete eyes. They have actually Photoshopped her eyes just slightly that it feels a bit off. They're a little bit larger. And humans are attracted especially to faces but especially the eyes. And so using that little slight bit of body dysmorphism actually is going to attract more attention and more intrigue around your thumbnail. And so they've done a really good job of that. I mentioned earlier too that her sucking up the worms is a half-completed action. We haven't done this study, but there have been psychological studies done on this subject where a half-completed action in a thumbnail is actually more clickable than a fully-completed action. So if I'm a basketball player, or if I'm doing a basketball channel, a player holding the ball or dribbling is not as exciting as someone mid-shot. A shot that's being completed with the ball going through the hoop is not as attractive as the ball right about to hit the hoop because we as humans are uncomfortable with incomplete action. We want to see how that story ends and so we click on that video to see how it all plays out. So again, action and half-completed action are the things that are going to really help optimize your content and get people excited to complete their journey and want to figure out how that video ends. So that's a bit about thumbnails. Let's move on to titles. So titles, obviously, the other gateway to any sort of video. And I think a lot of people kind of understand these basic principles, but again, questions. That whole idea of creating a story of kind of this sort of friction with your viewer that they want to have answered for them by watching your video. Who, what, how, why? Can you see a superhero? Can Sonic survive his own speed? How much do YouTubers make? Is it an interesting question? But if you take that, YouTubers make this much. YouTuber salaries exposed. That one's pretty good. But framing it in the form of a question actually is going to be better than if you just framed it as a statement. So you can frame a title in any number of ways, but framing it as a question is actually going to give you overall higher clicking rates. When it comes to numbers, you see top 10 lists across the board, right? Like that was a huge trend a couple of years ago. 10 is actually the strongest number when it comes to human psychology. People like the rounded completeness of 10, and so they're actually much more likely to click on videos with 10 than, say, 11 or 9. 10, the second place number is 5, actually. So if you're, you know, 5 facts and things like that. Not only... Again, it's a weird just kind of quirk in human psychology, but the number... 10 is number 1, then 5 is the second most clickable number. And then after that, odd numbers perform better than even numbers. Odd numbers for whatever reason. I don't know exactly what the psychological phenomenon is, but I've seen this in study after study after study that I've read on this topic, that odd numbers are just more attractive to viewers, to individuals, to just the human psyche. And then superlatives. I think we all know this. The best cupcake or the second smallest person in the world, everyone wants to see the most extreme, the biggest, the best, the worst. So framing your things as superlatives are immediately queuing in like, if I as a human have a limited amount of watch time, if I as a human have a limited amount of, you know, ability to watch this content, I want to engage with the stuff that is going to give me the biggest bang for my buck, and those are the superlative videos. Next up, we're addressing the audience. I'm going to actually fly through this because I'm running short on time. I've been rambling too much. But when you address the audience, we all hear that idea of all the YouTube creators introducing themselves. They all just kind of have their routines. They all say, hey, we all have our greetings, our signature sign-ins. This is all because humans love habits. They love rituals. They like feeling in a familiar space. They're good for, you know, the merch that you can stick it on a t-shirt, but you're actually playing into the human need to understand what's going on, to be in a safe space. And so starting your video with a ritual, ending your video with a ritual, having catchphrases that people recognize, all of that is allowing the viewer to become comfortable in your content, see it as a safe space, see it as a warm and accepting community that they are a part of. You're building that community around you and that, in turn, is going to lead to longer, long-term loyalty and viewership on your content. Set those expectations. Also, don't give people a chance to click away. So, again, to that idea of priming, to that idea of setting expectations, to that idea of people not wanting to click on black boxes on YouTube, when I click on a video, I want to see the topic of that video that was promised to me. And so, variety shows, if you're a tonight show, if you're a late night show, if there are multiple topics in the same 10-minute, 15-minute video, don't. See if there's a way that you can actually chop it up by the segment, because retention time data shows time and time and time again that anytime you switch to a different segment, anytime you switch to a different format, to a different topic, anytime you give your audience an opportunity of, like, this arc of the content is done, now we're moving into something else. You lose a significant amount of your audience and your retention time suffers as a result. And on YouTube, a platform that rewards you for higher amounts of retention time, higher amounts of watch minutes, you don't want to lose those precious viewer minutes. So making sure that you're not switching in between topics is very, very important, and ideally you're kind of... You've seen that evolution actually happen with the late night shows, where when they first started on YouTube, they were uploading huge chunks of the show. They were re-uploading the show. They were doing things like, and the game segment with the same celebrity. But I'm assuming... I haven't seen their data, and I wasn't in the rooms when those decisions were made, but you've seen them gravitate more towards here's the single segment, here's the game. Even though they're all united by the same personality, the same celebrity guest or whatever, they've separated out because retention time stay higher, they stay more engaged, and overall your watch minutes and retention percentage increase accordingly. And then the other thing we end off on the content and how to keep people engaged. Another fun fact here is speaking faster on YouTube is important. So in a typical conversation, you're speaking in about 100 to 150 words per minute. On YouTube, on the slow end, you're talking about 200 words per minute. And that's for some slower educational content. On average, it's ranging more in the 250 words per minute range. And to give you even more context to that, an auctioneer is speaking at 400 words per minute. So, you know, you're getting up into auction your territory there. So that's kind of the general flow of content. Vloggers speak faster than gamers. Different content verticals have different cadences, but in general all of them are 200 words per minute and up. And then lastly, end screens. This is the time to ask your audience what to do next. And here you're driven by this is one of the most important parts of your video, and here you're driven by two primary metrics, or two primary psychological phenomenon, which is decision fatigue and habituation. So decision fatigue is I as a human only have a limited amount of ability to decide what I want to do next. And I want that decision to be made easy for me. And so here we see an example of an end screen for my buddy Grant Thompson which we've talked to him about and he's since corrected, where there's too much happening on screen. He's asking, here's one video for you to click. Here's another video for you to click. I'm talking to you in the middle box. Here's a subscribe box and here's, you know, a third thing for you to click. There's too much happening on screen, and so I as the viewer don't know what to pick. I'm overwhelmed with choice, and when you overwhelm me with choice, I'm going to choose nothing. I'm going to choose nothing. But if you tell me, hey, click on, you liked this video? Click on this video because it's largely the same. Like, and here it is and it's the only thing. And there's a starburst behind it and I'm pointing to it. Now all of those things add up together to drive my attention in that way and I absolutely must click that video. You see click-through rates when you do a focus called action skyrocket. I say this all the time, but I say like, if I could design you an end card that gets you 0% click-through rate or 45% click-through rate and those are ultimately the differences here. Knowing what you're driving them to, driving them to a similar piece of content making sure that you convince them that hey, this is, you liked this piece of content this piece of content you will also like for X, Y and Z reasons here it is, click it right now and all of my focus is a tune to it and I click it. Again, remember, people don't like clicking black boxes, they want to know what they're getting into. Habituation, make sure you like, comment, and subscribe. It's a catchphrase on YouTube, no one cares. Again, drive the awareness by focusing on one thing and changing up the way that you say it. Like, comment, and subscribe people are habituated to it, they're used to it, it is meaningless to them now, they tune it out, so mix it up a little bit with like punch that subscribe button, lick my face to watch the newest video, things like that. The other interesting thing about the Grand Thompson example, he's talking in the middle frame, we've done tests with mouse scroll overs actually, where people will want to click the person who's talking to them even though that isn't an actionable part of the video, interestingly enough. And then, this is kind of reinforcing what we just said, and then since I'm out of time, this is the last thing and this is the last thing anyway. Commitmentphobia, YouTube makes it available to you to click on a playlist of 105 videos, 111 videos, watch the entire series, don't do that. So even though I, as a viewer know, I'm clicking into one video, I, even though it's one video, the fact that it says it's 100 videos, actually scares off viewers and those end screens, those end screen calls to action, actually get like 0% click through rate, because people don't want to commit to that. Even though they're not, even though they subconsciously know I'm not actually going to be watching 105 videos, they won't click on it because it feels like too much of a commitment. So anyway, long story short, thank you all for your time today. Remember, when you're optimizing your content, the single most important thing, yes think about the algorithm, yes think about playing nicely with all of these platforms and the way they share and distribute your content, but also make sure that you're not just playing to those neural networks, but you're playing to these ones. Thank you all so much. Enjoy big time.