and goal. The surface features viewers see are:
Browse Features: Homepage and Subscription
Suggested
Trending
Notification
Search
Each of these features runs separate algorithms trying to be optimized for a higher hit rate, and they all feed into the YouTube AI. They have separate hit rates to determine what actually works for users in each particular system. Hit rate means how often viewers are able to find what they actually want to watch. Have you ever heard of a fisherman getting a “hit?” It's when a fish takes the bait. Imagine that you are the fisherman who has tossed his video into the water. Potential viewers are the fish swimming by your “bait.” Maybe 10 fish take a look at the bait and swim on by because it's not the brand of bait that they like. But along comes a fish who says, “That looks good,” and he bites. Say you toss this line 10 times, and while 100 fish swam past, 10 took the bait. There's your hit rate. This hit rate is so important to each system in the AI. The algorithms are very sensitive to user behavior and the metadata on each traffic source so that they know how to change to increase the hit rate.
Additionally, YouTube is constantly running experiments—several thousand a year—and they implement about 1 in 10 changes as they go, so this translates to hundreds of changes being implemented annually. These changes help the system get smarter, and smarter means better at feeding viewers what they will watch.
Browse: Homepage
YouTube's Homepage has changed over time. Users no longer have to type a query in Search or to put in the work to navigate. The Homepage used to be where users saw only video recommendations of channels they had subscribed to. Now the Homepage has a personalized recommendation feed based on that user's history over time.
As long as a user is logged in when using YouTube, the algorithms can keep track of what videos that particular user has watched in the past so it can make better suggestions for videos that user is likely to watch, even if they haven't watched those videos in the past. This seems counterintuitive. It seems like the algorithm could be more successful suggesting videos it knows a user has already watched and liked, but actually the opposite is true. Keeping suggestions fresh actually gets users to stick around on the platform longer because they aren't getting bored with the same old stuff.
How YouTube does this is by breaking down the Homepage into two categories: familiar and discovered. It shows users familiar content from places the viewer has gone to before. These suggestions could include trending or recent videos from a channel a viewer has already watched. The discovery side includes videos or channels that users with similar viewing patterns have watched and liked. YouTube has found that these combined strategies are keeping viewers better engaged. If you want to get on the Homepage—and this definitely should be your goal—learn the triggers that get videos there. Try to improve your click‐through rate and audience retention, because these will help you get pushed out to a more general audience.
Browse: Subscription
This one doesn't need a lot of explanation. The Subscription section pulls content from channels you've already subscribed to. It will suggest new videos from your subscribed channels, especially new videos with similar content to what you've consumed before. For example, you watched a prank video or two from a channel you have subscribed to, so the AI pulls that channel's newest prank into your Subscription feed.
Suggested
Another place besides the Homepage creators should be focusing on is Suggested feed, including the “Up Next” video. These are the suggestions below (on mobile) or to the right (on desktop) of the video that is being watched. This is a powerful place to be! Viewers stick around when this feature is working really well, and sticky viewers are YouTube's goal. So if you can use the triggers to get your video in the Suggested feed, you're exactly where you need to be.
What are the triggers? First, make sure you have created a strong relationship among the data in your own content. This means that if the metadata connects among your videos, the algorithm plugs your videos into the Up Next feed, and your video's likelihood of being watched skyrockets. Metadata includes title, keywords, description, and the content itself. We talk in depth about metadata in Part III of the book. The viewers' behaviors trigger which videos get put in the Suggested feed as well, so when your content keeps viewers watching instead of bouncing, your content is more likely to be recommended here.
Other things the AI looks for in this feature include the “rabbit hole” type and the “watch something else” type. Rabbit hole explains itself pretty well. It's the type of videos that are similar in one way or another that keeps the viewer following a specific path. These include:
Videos from the same channel
Videos and channels that are similar to the one playing
Videos that other people watched after watching the current video
The watch something else type (also self‐explanatory) exists because viewers eventually tire of watching videos with similarities, and they need something entirely different if they are going to stick around. This isn't a random selection; it's still a personalized recommendation based on their past behavior. This recommendation comes when the AI has a stored history of what that viewer has watched over time.
Trending
I like to call this the geo‐specific water cooler. A common misconception is that “trending” is synonymous with “popular,” but it's not. Trending topics are broad topics that people are talking about right now all across the Internet. It is what's currently happening in the news, on social media, websites, blogs, and elsewhere. By “geo‐specific” I mean that even the Internet has geographical regions, both in location and in authority. Websites with more authority than others will have topics that trend better because YouTube knows more eyes will be on them, so it pulls those topics into their own Trending section. They realize it's what people are seeing elsewhere so they're more likely to click on that topic when they are on YouTube as well. By location, groups that live in the same area often are interested in the same things. For example, something that would trend well in LA likely would fall flat in the Midwest.
Notification
When someone subscribes to your channel they can be notified by YouTube when you upload a new video, but only if that subscriber has also clicked on the bell button. The notification comes through on the subscriber's YouTube app or via email.
Search
Another straightforward feature is the Search. Users type in a keyword or phrase as a query to find what they want to watch. The Search feature displays videos related to that query. The algorithm narrows down the results based on the metadata and the video generated by the creator, and it also looks closely at past data from people searching similar queries, and how they responded to those videos. A lot of people think they just need to do SEO to make their video go to the top, but there is a “freshness” feature that will pull new videos into these results. Take a look at what's trending and create content with the right connecting metadata. Know especially what's trending in your niche. If your video performs well, the algorithm keeps it in the Search results, but if it
