Facebook Watch is a video service by Facebook where users can view videos posted by individual users, media companies, and creators. The videos that show up on a user’s Watch feed are determined by several factors, including the user’s interests, who they follow, video popularity, paid promotions, and Facebook’s algorithms.
A User’s Interests and Follows
One of the main factors that determines which videos show up on a user’s Facebook Watch feed is their own interests and the pages/people they follow. When a user first signs up for Facebook, they are prompted to list some of their hobbies and interests. This allows Facebook to tailor video recommendations to content the user is likely to enjoy watching.
In addition, the videos posted by pages and individual users that someone follows will appear higher up in their Watch feed. If you follow a lot of food pages or friends who post cooking videos, your Watch feed will surface more cooking-related content. By following certain accounts and interests, you have control over shaping the types of videos Facebook recommends you watch.
Related Interests
Facebook will also suggest videos based on interests related to the pages/people you already follow. For example, if you follow several yoga pages, Facebook may recommend videos from meditation or fitness pages. This helps expose users to new videos and accounts that align with their likes.
Friends’ Likes and Shares
In addition, Facebook will factor in what your friends have liked, shared, and commented on. If a cooking video was popular among your network of friends, it is more likely to end up on your Watch feed even if you don’t follow any cooking accounts. Friends’ engagement with certain videos signals to Facebook that it is trending or viral content you may be interested in.
Video Popularity
In general, the more popular a video is on Facebook, the more likely it will be surface high up on users’ Watch feeds. Popular signals Facebook looks for include:
- High view count
- Lots of reactions, comments, and shares
- Frequently shared or embedded on external sites
- Watch time and completion rates
Videos that are going viral and have high engagement will get prioritized by Facebook’s algorithm so they get viewed by even more people. This creates a snowball effect for the top trending videos.
Recirculation
Additionally, once a video goes somewhat viral, Facebook will start heavily recommending it to new audiences who have not seen it yet. So you may start noticing the same videos being recirculated if they are performing well. Facebook wants to ensure viral videos reach as many viewers as possible.
Paid Promotion
Pages and creators can pay Facebook to promote their videos to certain demographics of users. When a video has been boosted through Facebook or Instagram ads, it will appear more frequently on users’ Watch feeds. Pages want their videos to reach and engage with new viewers, so they invest in paid promotion.
Here are some common ways videos get promoted on Facebook:
- Boosted Facebook posts featuring the video
- Video ads running on Facebook/Instagram
- Promoted posts in users’ News Feeds
- Sponsoring the video to appear at the top of Watch feeds
The more money put behind a video ad, the more users it will be shown to. Pages with big budgets can get their videos seen by tens of millions of people through paid promotion.
Facebook’s Algorithms
The final key factor that determines which videos appear in Watch feeds is Facebook’s proprietary algorithms. Based on the myriad of signals and user data available, their AI models customize each user’s individual video recommendations.
Machine Learning
Facebook uses advanced machine learning techniques to get better at predicting which videos each user is likely to watch, share, and engage with. By analyzing user behavior patterns, Facebook’s AI pinpoints videos that align with someone’s taste.
Objective
The overarching objective of Facebook’s algorithm is to surface videos that will keep users watching for longer periods. The more time spent watching, the more video ads Facebook can show. This means videos that Facebook predicts will be interesting, relevant, and engaging for each user get prioritized.
Optimization
Facebook is constantly monitoring how videos in Watch feeds perform and optimizing based on the data. If a certain category of videos continually leads to higher average watch times, the algorithm will serve up more of that type of content. The goal is to optimize Watch feeds to be as addictive as possible.
New vs Established Accounts
There is also some variance in how Facebook’s algorithm treats newer accounts compared to established accounts with more data. For newer users, Facebook may rely more heavily on broader signals like overall video popularity and interests listed in the user’s profile. For established accounts with more data, the algorithm can better customize recommendations based on years of viewing behavior.
Cold-Start Problem
This “cold-start problem” for newer accounts means it may take some time before Facebook can pinpoint that user’s unique tastes. As the algorithm gains more familiarity with what each user watches, likes, and shares over many sessions, it gets much better at surfacing personalized videos tailored to individuals.
Factors That Don’t Matter
Some things that do NOT impact what videos show in Facebook Watch feeds include:
- The age of the video – Older viral videos can still appear frequently if they align with user interests
- Alphabetic or chronological order – Video order is based on relevance, not recency or name
- Device used – The algorithm aims to provide a consistent experience across mobile, desktop, etc.
The key driving factors are users’ tastes, video engagement metrics, and Facebook’s optimization of recommendations to keep people watching longer.
Conclusion
In summary, the videos that appear on users’ Facebook Watch feeds are determined by a mix of factors:
- Pages and friends the user follows
- The user’s interests and Facebook activity
- Engagement and popularity of videos
- Paid promotions by pages/creators
- Facebook’s machine learning algorithm
By optimizing Watch feeds to align with these signals, Facebook aims to surface personalized, relevant videos that will capture users’ attention and get them watching more content on the platform.