TikTokJump to a section
What are recommender systems? • How content is recommended across TikTok • Other factors that influence how TikTok recommends content • How you can influence what you see on TikTok
What are recommender systems?
TikTok's mission is to inspire creativity and bring joy. We have a wide variety of content and we want you to discover interesting and relevant videos. That’s why we use recommender systems to offer you a personalized experience. These systems suggest content based on your preferences as expressed through interactions on TikTok, such as following an account or liking a post.
When you first sign up for TikTok, we may invite you to select categories of interest, like pets or travel, which help us form your For You and LIVE feeds. If you don't select any categories of interest, we start by offering you a feed of recent videos that are popular with others on TikTok. When selecting your initial set of videos, the recommender system looks for popular videos that are appropriate for a broad audience and influenced by your location and language settings. In addition to categories, we'll suggest popular creators for you to follow. If you don't select any followers or have any friends on TikTok, then you may not see content in your Friends Tab or Following feed.
Once you start interacting on the app, your interactions act as signals that help the recommender system predict content you're more likely to be interested in as well as the content you might be less interested in and may prefer to skip. Your interactions across TikTok can impact how the system ranks and serves you content.
How content is recommended across TikTok
We recommend content based on several factors to predict how relevant and interesting content might be to a user. The three main factors are user interactions, content information, and user information.
Each recommender system selects from a large collection of eligible content and ranks them based on the system's prediction of how likely you’ll be interested in each one. In certain recommender systems, these predictions are also influenced by the interactions of other people on TikTok who appear to have similar interests. For example, if User A likes videos 1, 2, and 3 and User B likes videos 1, 2, 3, 4, and 5, the recommender system may predict User A will also like videos 4 and 5. While some may see the same content, features like the For You, Following, Friends, and LIVE feeds are unique to each person.
You can view content through several feeds and features, which we continually improve and expand over time. Below, we've provided some popular examples.
For You
The For You feed displays a stream of content just for you, so you can find content and creators you love.
Factors that may influence your For You feed include:
• User interactions: Content you like, share, comment on, and watch in full or skip, as well as accounts of followers that you follow back.
• Content information: Sounds, hashtags, number of views, and the country the content was published.
• User information: Device settings, language preference, location, time zone and day, and device type.
These factors help us predict how relevant and interesting content might be to a user. Different factors can play a larger or smaller role in what's recommended, and the importance—or weighting—of a factor can change over time. For many users, the time spent watching a specific video is generally weighted more heavily than other factors.
Following
The Following feed displays content from creators you follow. Each Following feed is unique—even when two users follow the same set of creators, each feed will rank and serve content in a way that is tailored to how you use TikTok.
Factors that may influence your Following feed include:
• User interactions: Visits to the user profiles you follow and the content you like, share, comment on, and watch in full or skip.
• Content information: Sounds, hashtags, number of video views, and the country the content was published.
• User information: Device settings, language preference, location, time zone and day, and device type.
These factors help us predict how relevant and interesting content might be to a user. Different factors can play a larger or smaller role in what's recommended, and the importance—or weighting—of a factor can change over time. For many users, the time spent watching specific videos is generally weighted more heavily than other factors.
Friends
The Friends Tab displays content from followers that you follow back (your friends), accounts you follow, and other suggested accounts. Even when two users follow the same set of creators, their Friends Tab will rank and serve content in a way that is unique to each person.
Factors that may influence the Friends Tab include:
• User interactions: Visits to the profiles of your friends or people you may know and the content you like, share, comment on, and watch in full or skip.
• Content information: Sounds, hashtags, number of video views, and the country the content was published.
• User information: Device settings, contacts (with your permission), language preference, location, time zone and day, and device type.
These factors help us predict how relevant and interesting content might be to a viewer. Different factors can play a larger or smaller role in what's recommended, and the importance—or weighting—of a factor can change over time. For many users, likes and comments on a video are generally weighted more heavily than other factors.
LIVE
The LIVE feed displays a stream of LIVE videos that allows you to find content and creators you love. You can view LIVE videos by selecting the LIVE button at the top of the For You feed, or you may see recommended LIVE videos in your For You Feed.
Factors that may influence your LIVE feed include:
• User interactions: LIVE videos you like and comment on, creators you follow, watch time, and Gifts sent.
• Content information: Hashtags, number of LIVE views, number of likes, number of Gifts sent, number of followers of that creator, and the country in which the LIVE is taking place.
• User information: Device settings, language preference, location, time zone and day, and device type.
These factors help us predict how relevant and interesting content might be to a user. Different factors can play a larger or smaller role in what's recommended, and the importance—or weighting—of a factor can change over time. For many users, the time spent watching a specific LIVE is generally weighted more heavily than other factors.
Search
Our search feature allows you to search for content on TikTok. It also recommends search terms for you to discover content.
Factors that may influence your search results include:
• User interactions: Your past search behavior and the content you like, share, comment on, and watch in full or skip.
• Content information: How relevant the content is to your search, hashtags, and the sound used.
• User information: Device settings, language preference, and location.
These factors help us predict how relevant and interesting content might be to a viewer. Different factors can play a larger or smaller role in what's recommended, and the importance—or weighting—of a factor can change over time. For many users, how relevant the content is to the specific term entered is generally weighted more heavily than other factors.
Comments
We recommend comments on posts that we think you'll be interested in and to build connections within the TikTok community.
Factors that may influence what comments we show you:
• User interactions: Comments you like, reply to, and if you follow the video creator or author of the comment.
• Content information: Number of likes and the language of the comment.
• User information: Your location and language preference, and device settings.
Different factors can play a larger or smaller role in what's recommended, and the importance—or weighting—of a factor can change over time. For many users, the language of the comment and how many likes it has are generally weighted more heavily than other factors.
Notifications
We'll send you notifications to recommend content you might like, people you might know, or search terms you might be interested in.
Factors that may influence what notifications we send you:
• User interactions: How active you've been on TikTok and your previous interactions with notifications.
• Content information: Creator of the content, hashtags, number of views, and the country the content was published.
• User information: Your location, language preference, and device type.
Different factors can play a larger or smaller role in what's recommended, and the importance—or weighting—of a factor can change over time. For many users, the number of notifications you receive and open is generally weighted more heavily than other factors.
Account recommendations
On TikTok, you can connect with people you may know and creators that might interest you, based on who you're already engaging with.
Factors that may influence your account recommendations include:
• User interactions: Your activity on TikTok, such as the types of accounts you follow, like, or share content with.
• Content information: Number of followers and views an account has, account location, and the likelihood that you know another user, for example if they're in your contacts (with your permission).
• User information: Device settings, language preference, and location.
Different factors can play a larger or smaller role in what's recommended, and the importance—or weighting—of a factor can change over time. For many users, the number of mutual connections between two accounts is generally weighted more heavily than other factors.
Other factors that influence how TikTok recommends content
Diversifying recommendations
When you view content in your Friends Tab, For You, Following, and LIVE feeds, our goal is to find a balance between suggesting content that's relevant to you while also allowing you to discover new and diverse content and creators, and experience new perspectives and ideas. That's why you may come across content in your For You and LIVE feeds that doesn't appear to be relevant to your expressed interests. We typically don't recommend content you've already seen and we'll encourage you to explore different categories of content and creators through recommendations.
Similarly with your Friends Tab and Following feed, you may see multiple videos from the same creator, but we generally won't recommend two videos in a row from that creator. If you follow only them, then we may recommend multiple videos in a row. If you run out of videos to watch on your Friends Tab or Following feed, then we'll suggest for you to explore videos on your For You feed.
Safeguarding your experience
Our recommender systems are designed with safety as a key consideration and we remove any content identified as violating our Community Guidelines.
Our safety team takes additional precautions to review content as they rise in popularity to reduce the likelihood of recommending content that may not be suitable for everyone on TikTok. For example, we strive to not recommend—or to limit the recommendation of—certain categories of content that may not be suitable for a general audience, even if that content is not removed from TikTok altogether. We continually review LIVE videos to interrupt and stop LIVE content that violates our Community Guidelines. We may also restrict creators from going LIVE who violate our Community Guidelines. Users under the age of 18 are not permitted to host a LIVE or to send Gifts, and are restricted from watching certain LIVE videos. Learn more about age-restricted content on LIVE.
You can learn more about our standards on content eligibility in our Community Guidelines.
How you can influence what you see on TikTok
For You
In addition to selecting categories of interest when you sign up for TikTok, you can use the following features to help shape your For You feed:
• Not interested: If you don't enjoy a specific piece of content you can let us know by sharing feedback that you're not interested in and we'll show you fewer content like it.
• Filter video keywords: Filter out specific words or hashtags from the content recommended to you in your For You and Following feeds.
If the recommendations on your For You feed don't feel relevant to you anymore, or you're looking for more variety in your feed, you can refresh your For You feed to view a new set of popular content, as if you've just signed up for TikTok.
Following
Use the following features to help shape your following feed:
• Unfollow: Unfollowing a creator can directly remove the creator's posts in the Following feed.
• Filter video keywords: Filter out specific words or hashtags from the content preferences section in your settings to stop seeing content with those keywords in your For You and Following feeds.
Friends
Use the following features to help shape your Friends Tab:
• Unfollow: Unfollowing the creator or friend could remove the posts from that account in the Friends Tab.
• Not interested: If you don't enjoy recommended content from people you may know based on suggested accounts, you can let us know by sharing feedback that you're not interested and we'll show you fewer content from that creator.
LIVE
In addition to selecting categories of interest when you sign up for TikTok, you can use the following features to help shape your LIVE feed:
• Not interested: If you don't enjoy a specific piece of LIVE content, you can let us know by sharing feedback that you're not interested and we'll show you fewer content from that creator.
Comments
Let us know if you're not interested in a comment by disliking it within the comments section of a post.
Notifications
Manage notifications in your settings to choose what type of notifications you'd like to receive for suggested posts. Turn all notifications off or choose which interactions you'd like to receive notifications for.
Account recommendations
Use the following features to help shape your account recommendations:
• Unsync your contacts: Remove previously synced contacts or turn off synced contacts in your privacy settings.
• Remove suggested accounts: When we recommend accounts for you to follow, you can remove it from the suggested list to let us know you're not interested.