Rachel Alves shares a video on “How video recommendation system of YouTube” works to help creators understand how the system works
Recently, YouTube has published a new overview on how Search and Discovery work on the platform. This was done in order to help the creators of the platform build an audience and reach more viewers with their youtube videos. This YouTube overview was presented by Rachel Alves, a product manager for discovery at YouTube.
She has shared various insights on how the platform’s video recommendation system actually works. The presentation she presents during this overview is most likely to help people understand how the system works and how videos are shown to prospective viewers.
The key aim as mentioned was: Help each viewer find videos they want to watch and maximize long-term viewer satisfaction. The key aims point to say that the platform ensures to help viewers with what they want to see and ensure they are satisfied with it.
YouTube’s goals have always been the same, what it optimizes now has changed over time. “If we go back to 2011, what we optimized for was clicks and views, but that’s not that great of a metric, because it may indirectly incentivize clickbait or sensational titles or thumbnails that get people in to watch a video, but doesn’t make them very satisfied or happy”. – Rachel Alves.
The feedback YouTube got in its early days about its algorithmically defined feed was that viewers’ feeds were filled with “off-putting” videos. Therefore, in 2012 YouTube switched its focus to watch time as its a key metric. However, she says this is not a perfect metric either, and thus, YouTube has been seeking to optimize the viewer’s feed according to their level of satisfaction. YouTube makes sure to do this by asking users through surveys and reducing the streaming of violative content that increases the risk of spreading misinformation.
The overall concept of YouTube’s recommendation system is to give viewers utmost satisfaction with the content. To make sure to get on the correct line, YouTube rolls out surveys to million users per month. Not only this, but YouTube also sends direct feedback signals to the users. YouTube uses signals like When people tap/click on the ‘Not interested’ option in the individual video menu, Likes and dislikes on videos, and Shares of videos.
Creators often want to know how they can optimize for each element in YouTube, to which the hard answer is “you just can’t”. Alves says, “You can’t optimize for a traffic source, you can only optimize for people or viewers.” Generally the content that is being shown on the feed of a viewer is based on the performance of the particular video. It’s not just that, it is also based on the personal interest of the viewer. So, basically, the algorithm suggests videos on a homepage that are not only doing well but are also of the viewer’s interest.
Though you can’t optimize for each element on the home page, you can surely look at your content. You have to think on the lines that what would a viewer want to see in the content and if they are being recommended to watch my video, will they watch it. Talking about “suggested” videos that are recommended after the video viewers are currently watching are also based on viewers’ relevant interest.
You can’t optimize for the homepage but you can definitely optimize for suggested. The most effective tactic seen is that creators used to maximize their appearances in people’s ‘Suggested’ listings is to develop a video series or create topically related videos that lead on from one another. This is more likely the videos that the viewers often tend to watch in one session, one after another. She also recommends using a consistent title and thumbnail style that makes it easy for the viewer to find a particular creator’s videos.
You can imagine when a viewer is looking at everything that they could choose to watch next, there are a lot of options there, and if you have really strong, identifiable branding, that’s consistent, it’s really easy to pick out which videos are from your channel, and it just makes that decision all the quicker for viewers. CTA buttons to ‘Watch more’, as well as playlists and end screens are also effective tools in order to encourage viewers to keep watching your content.
Sometimes creators have a perception that the recommendation system pushes out or promotes videos to viewers, when in reality, the system is designed to work the opposite way, where a viewer visits youtube.com, and then a recommendation system pulls in and then ranks the best candidate for that viewer, depending on the page that they’re on.