The Recommended for you feature uses artificial intelligence to categorize your video library and your members' viewing habits, then surfaces personalized video suggestions directly in the catalog.
This creates a Netflix-style discovery experience that helps members find relevant content without browsing through every category manually.
The Related videos show a more intelligent list of related videos with similar characteristics.
This lets members engage with more relevant content and find their next video to watch faster.
đ NOTE: The Video Recommendations feature is currently in an early beta access for eligible Uscreen stores. The Content > Recommendations page only appears in your admin dashboard if your store is part of the current beta cohort. If you donât see it, your store has not yet been enabled.
How It Works
The AI Video Recommendations feature has two main components:
1. "Recommended for you" Catalog Row
A new dynamic category row appears in your catalog for signed-in members who have some viewing history, showing a personalized selection of videos based on their viewing history and preferences. The row is placed prominently at the top of the catalog to maximize content discovery.
2. Intelligent Related Videos
On individual video pages, the related videos section is enhanced with intelligent suggestions. Instead of only showing videos from the same category, the system shows videos that are genuinely similar in content, style, and topic, helping members discover relevant content across your entire library.
Catalog Row Placement
When enabled, the Recommended for you row appears in the following order on the catalog page:
Featured Category (hero banner)
Recommended for You (NEW)
Continue Watching
My Library
Favorites
Admin-defined categories
âšď¸ INFO: The Recommended for you category row only appears for signed-in members who have some viewing history. Anonymous visitors and brand-new members without watch history will not see this row, similar to how Continue Watching and My Library behave.
What Powers the Recommendations
Behind the scenes, the system works in multiple stages:
AI Video Tagging: Each video in your library is automatically categorized. The system processes video title and description metadata, transcript from subtitles, and a small sample of video thumbnail screenshots to generate detailed tags describing topics, difficulty level, equipment, workout style, intensity, duration, body focus, and more. AI does not see your raw video file. Your videos do not train AI models.
Taste Profiles: As members watch videos, the system builds a taste profile for each member based on the types of content they watch and complete.
Smart Matching: The recommendation engine compares each member's taste profile against your tagged video library to surface the most relevant unwatched content.
đĄ TIP: The more content a member watches, the better their recommendations become. Encourage new members to explore a few videos to help the system learn their preferences.
Viewing & Managing Recommendations
If your store is eligible for Video Recommendations, youâll see Recommendations under Content in your admin dashboard.
The Content > Recommendations page shows the current status of AI-powered recommendations for your store, including:
Videos tagged: The number of videos auto-tagged by AI and ready to power recommendations.
Members profiled: The number of members with a taste profile based on their watch history.
Active model: The recommendation model currently powering recommendations.
From this page, you can also manage the recommendation surfaces available to your members:
Recommended for you: A personalized row in each memberâs catalog, built from what theyâve watched.
Related videos: AI-selected related videos are shown on each video page.
You can also use the links on this page to:
Manage which videos are recommended: Review tagged videos and exclude any videos you donât want recommended.
View member taste profiles: See what the AI has learned about each member from their watch history.
Managing Video Recommendations
You can control which videos appear in recommendation surfaces, either per video or in bulk from the videos list.
âšď¸ INFO: This setting is available to users with the Owner, Admin, or Content Manager role.
Per-Video Control
Navigate to Content > Videos
Find and open the video you want to manage
Scroll down to the Recommendations card on the video details page
Select either:
Show in recommendations: the video will appear in the "Recommended for You" catalog row and the "Related videos" section under other videos
Hide from recommendations: the video will be excluded from all recommendation surfaces
Bulk Actions
Navigate to Content
Select the videos you want to update by ticking the checkboxes beside them
Open the bulk actions menu
Select Show in recommendations or Hide from recommendations
FAQs
Why donât I see Recommendations under Content?
Why donât I see Recommendations under Content?
The Recommendations page only appears for stores in the current eligible beta cohort. If you donât see it, your store has not yet been enabled for Video Recommendations.
How many videos appear in the Recommended for You row?
How many videos appear in the Recommended for You row?
The row displays a selection of up to 8 recommended videos. Only unwatched videos get recommended.
Can I manually curate what appears in the Recommended for You row?
Can I manually curate what appears in the Recommended for You row?
Yes. While the row is powered by AI based on each member's viewing behavior, you can control which videos are eligible to appear in recommendations. See Managing Video Recommendations for details.
What does Members profiled mean?
What does Members profiled mean?
Members profiled stat shows how many members have a taste profile based on their watch history. Members need some viewing activity before the system can personalize recommendations for them.
Whatâs the difference between Recommended for you and Related videos?
Whatâs the difference between Recommended for you and Related videos?
Recommended for you is a personalized catalog row based on each memberâs watch history. Related videos uses AI to choose relevant videos shown on individual video pages.
Will members with no watch history see recommendations?
Will members with no watch history see recommendations?
Members need some viewing history for the system to build a taste profile. New members without watch history will not see the Recommended for you row until they've watched some content.
Does this replace the existing Related Videos section?
Does this replace the existing Related Videos section?
No. The Related Videos setting on the Recommendations page enhances the existing Related Videos section with AI-selected matches. When enabled, related videos on video pages are chosen based on content similarity instead of relying only on shared categories.
Does this work on mobile and TV apps?
Does this work on mobile and TV apps?
Available on the web today for participating beta customers. Available on mobile app version 3.33 or later. TV app support is not yet available, but will be soon.
Will this affect my catalog's performance or loading speed?
Will this affect my catalog's performance or loading speed?
No. The recommendations are computed in the background and served efficiently. There is no noticeable impact on page load times.
Are my videos being used to train AI?
Are my videos being used to train AI?
No, your videos are not training AI models. We are using two OpenAI models to categorize and recommend videos. We selected these models specifically, not used to train AI, and to keep your content safe.
đŹ As this feature is in early beta access, the behavior and availability described here may evolve as we gather feedback and expand the rollout.