There are a few main reasons why you may see people who don’t have any mutual friends show up in your Facebook “People You May Know” recommendations:
1. You have overlapping networks
Even if you don’t have any direct mutual friends with someone, you may still have overlapping social networks. For example, maybe you went to the same school or workplace, or live in the same area. Facebook’s algorithm looks at these types of connections to find people who are likely to be within your extended social network.
2. You have similar interests and online behavior
Facebook tracks the pages and content you engage with on its platform as well as other sites and apps. If you and another person like, comment on, or otherwise interact with similar pages and content, Facebook’s algorithm may suggest you connect.
3. The other person looked up your profile or contacts
If someone who doesn’t know you looks up your Facebook profile or looks at the profiles of people in your friend list, Facebook may suggest connecting you both since there appears to be some interest there.
4. Shared contacts or activities outside Facebook
Facebook can access your contact lists and activities on other platforms like Instagram if you give it permission. If you and another person have each other’s contact info or interact on another social platform, Facebook may suggest you connect.
How Facebook’s People You May Know Algorithm Works
Facebook uses a complex proprietary algorithm to generate People You May Know suggestions. While the exact details are not public, here are some of the key factors we know the algorithm considers:
Mutual friends
The number of mutual friends you share with someone is a strong signal. In general, the more friends you have in common, the more likely Facebook will suggest connecting.
Mutual interests and groups
If you and another person are both members of the same Facebook groups or like/follow the same pages, you are more likely to be suggested to each other.
Geographic proximity
People who live in the same area, visited the same places, or recently traveled to the same location may be suggested.
Education and workplace networks
Going to the same school or working for the same company creates a connection, even if you don’t have mutual friends.
Contacts and external platforms
If you give Facebook permission to access your contacts, calls, or activities on other platforms like Instagram, it can match you with people in your contact lists or who you interact with outside of Facebook.
Profile views and searches
If someone views your profile or searches for your name, Facebook may suggest you connect.
Keywords and tags
Shared keywords, facial recognition tags, and other data points from posts and metadata can also influence suggestions.
Past interactions
If you’ve previously interacted with someone on Facebook through commenting, tagging, reacting, messaging, etc. there’s a stronger chance you’ll be reconnected.
Controlling Facebook’s People You May Know Recommendations
If you want to limit who Facebook recommends in People You May Know, here are some steps you can take:
Adjust privacy settings
Go to Settings > Privacy and adjust who can see your friend list, posts, contacts, and other profile info. Making this info more private reduces material Facebook’s algorithm can use.
Limit profile visibility
In Settings > Privacy > How people can find and contact you, you can hide your profile from searches, disable profile suggestions, and limit who can send you friend requests.
Remove imported contacts
Go to Settings > Your Facebook Information > Downloading Your Information and remove any contacts you previously uploaded.
Turn off location services
Disabling location services for Facebook in your device settings prevents recommendations based on shared locations.
Delete old posts and limit public interactions
This gives Facebook’s algorithm less data to assess your interests and networks.
Actively manage recommendations
Click the X icon next to any recommendations to signal you don’t want to connect. You can also remove inadvertent profile views that may produce suggestions.
Pros and Cons of Facebook’s People You May Know
Facebook’s People You May Know recommendations have some potential benefits but also come with risks:
Pros
- Reconnect with old classmates, colleagues, etc.
- Discover new friends with shared interests or networks
- Grow your social graph on Facebook
Cons
- Privacy concerns over data collection
- Unwanted contact from loose ties, strangers, or vague acquaintances
- Stalkers and abusers can more easily connect
- Feature enables tracking of offline activity
Best Practices for Managing Recommendations
Here are some tips to manage People You May Know suggestions:
- Customize privacy settings and limit profile visibility
- Avoid connecting with vague acquaintances or strangers
- Proactively delete contacts, posts, and interactions you don’t want assessed
- Disable location tracking and hide your friend list
- Decline unwanted recommendations using the X icon
The Future of Social Graph Expansion
Here are some possible ways Facebook could evolve People You May Know:
- Leverage AI facial recognition on photos and videos to suggest connections
- Incorporate biometric data points into matching algorithms
- Partner across social platforms to suggest users from other sites
- Analyze off-Facebook activity like calls, texts, emails
- Suggest extended network connections up to 3rd or 4th degree
While offering potential value, growth of the social graph also risks amplifying the spread of misinformation, polarization, and compromising privacy. Facebook will need to tread carefully as it expands its friend recommendation capabilities.
Conclusion
Facebook’s People You May Know uses sophisticated algorithms to suggest new connections based on mutual friends, interests, networks, location, profile views, and online and offline interactions. While this can help reconnect old ties, it also comes with privacy tradeoffs. Users should actively manage recommendations and customize privacy settings to limit unwanted contacts. Going forward, Facebook will need to balance the value of social graph expansion with risks like misinformation and stalking.