Members of YouTube’s recommendation system and analytics team have explained how different factors influence the recommendation algorithm of the video platform, such as external traffic, inactive subscribers or video performance.
In a video posted via the Creator Insider channel aimed at the creators of the platform, two members of the YouTube team have answered frequently asked questions from creators about the recommendation algorithm.
Team members say that if one or more creators’ videos underperform the rest, this doesn’t have to hurt their channel’s performance in the long run, as the recommendation algorithm will always be “following to the audience”.
In this sense, if a video attracts an audience and goes viral, it will be shown in the recommendations of users, regardless of the performance of previous videos from the same channel.
Likewise, the members of the company have explained that there is no limit on the number of videos from the same channel that can be recommended to users in a single day, so that creators can upload the number of videos they want and the YouTube’s recommendation algorithm will recommend them “as long as viewers are willing to see them.”
Although there is no limit for recommendations, there is a limit for notifications, since users only receive three new video notifications per channel every 24 hours.
For inactive subscribers, the company advises creators not to create a new channel if they have a lot of subscribers who don’t watch their videos. Even if a channel has inactive subscribers, the algorithm can still show new videos in the recommendations section, whereas if the creator decides to create a new channel, it is more difficult for the videos to be recommended to users.
External traffic is a factor that influences YouTube’s recommendation algorithm and can help a specific video appear in the recommendations section, as explained from the platform, although the time it will appear in that section will depend on how the users respond. users.
In addition, the members of the company have stressed that creators should not worry if they receive a lot of traffic from external web pages that make their videos have fewer ‘clicks’ and that the average duration of visits is less, since it does not have any impact in long-term success.
In this sense, the recommendation system learns how a particular video is performing when it is recommended to viewers.