Facebook’s ‘People You May Know’ feature suggests new friends and connections to users based on several factors. The algorithm looks at mutual friends, work and education information, location, groups and pages liked, and other signals to determine which people you may know on the platform.
Mutual Friends
The number one factor that influences People You May Know suggestions is mutual friends. If you and another person have several friends in common on Facebook, it’s highly likely the algorithm will recommend you connect. Having mutual friends indicates you may run in the same social circles and know each other offline.
The more mutual friends you share, the higher the chance that person will show up in your recommendations. For example, if you and John Doe have 50 mutual friends, Facebook may prominently suggest John Doe to you. But if you and Jane Smith only have 2 mutual friends, she may not appear or may be buried further down the list.
How Many Mutual Friends Are Required?
Facebook has not publicly shared the exact number of mutual friends required to trigger a recommendation. However, research indicates that having 5-10+ mutual friends significantly increases the likelihood someone will show up in People You May Know.
Having just 1 or 2 mutual friends is unlikely to be enough for Facebook to make a suggestion. The more mutual friends you have, the better chance the algorithm will connect you.
Importance of Mutual Friend Connections
In addition to number of mutual friends, Facebook also looks at how well-connected those mutual friends are to each of you. If your mutual friends are close friends with both you and the other person, that carries more weight than if they are casual acquaintances of both of you.
For example, if 5 of your closest high school friends are also close longtime friends of John Doe, Facebook may recommend him since those mutual connections are strong. But if your 5 mutual friends with Jane Smith are just distant work colleagues or acquaintances, the recommendation may not happen.
Workplace and Education
Another important factor is whether you and the other person work at the same company or went to the same school. Having workplace or education affiliations in common significantly increases the likelihood of a People You May Know recommendation.
If you and another Facebook user work at the same organization or went to the same college around the same time, you probably have some overlap in your social and professional networks. Even if you don’t have any mutual friends, the algorithm may suggest connecting based on shared work/education backgrounds.
School Network Recommendations
For education, Facebook not only looks at whether you went to the same college or university, but also considers:
- Whether you were enrolled at the same time
- Shared major, degree, or coursework
- Overlap in extracurricular activities
The more precise the education match, the more likely the recommendation. For example, if you and Jane Doe both graduated from the University of Florida in 2019 with marketing degrees, the chance she’ll be suggested is higher vs. simply having gone to the same large state school at different times.
Company Recommendations
For employment, common matching factors include:
- Working at the same company currently or in the past
- Being in similar roles or departments
- Having colleagues in common
Just listing that you work at a large company like Walmart doesn’t necessarily mean all employees will be suggested to each other. But sharing the same employer, office, team, manager, or professional circle increases the likelihood of recommendation.
Location
Where you live, work, and spend time also influences People You May Know. If you and another person live in the same city or are frequently in the same geographic area, you are more likely to run in similar social circles.
Facebook may pick up location signals from:
- Current city listed in your profile
- Location tags on your posts and photos
- Places you check into
- Local events you mark interest in or attend
If you and John Doe both live in Miami, regularly check into downtown Miami hotspots, and attend the same local events, the algorithm may recommend connecting based on your geographic proximity. Even if you have no mutual friends, the location match indicates you likely have overlapping networks and could know each other.
Custom Audiences From Location Data
Facebook also creates Custom Audiences for advertisers based on location data, for targeting ads and building lookalike audiences. People You May Know suggestions may factor in these Custom Audiences if you and another user appear in the same geographic audience segment frequently.
Pages and Groups
Sharing common interests and community affiliations also influences People You May Know suggestions. If you and another person like, join, follow, or engage with the same Facebook Pages or Groups, you may have similar tastes, views, identities, or causes.
For example, if you and John Doe both follow the same local community group, political party page, hobby interest page, or other niche pages, Facebook may suggest connecting due to your shared interests or views.
Liking or joining a Page essentially “links” you to other people who also like that Page. So if enough Pages in common emerge between you and another user, Facebook sees it as a relevant signal to recommend friendship.
Page Admin Connections
Being an Admin or Editor of the same Facebook Page can also trigger recommendations between Page admins. This is because Page admins often know each other through managing that joint property, even if they have no other obvious connections.
Events
RSVPs to local events can also influence People You May Know. When you and other people mark yourself as “Going” or “Interested” in the same events – especially niche community events – it indicates you may run in the same social circles.
For example, if you and Jane Doe are both attending an upcoming local 5K race event, neighborhood street fair, PTA meeting, or alumni networking night, you will likely interact in real life. So Facebook may suggest you connect online as well.
Importance of Niche Event Overlap
As with Pages, the more niche and specific the event, the stronger the signal. If you and another user are both attending a concert with 50,000 people, it doesn’t necessarily mean you know each other. But having interest in the same local community theatre play, club sports team, or hobby group points to shared bonds.
Shared Contacts
If users upload their phone contacts to Facebook, it can also inform recommendations. If you and John Doe have each other’s phone numbers stored in your offline address books, or have overlapping contacts, Facebook may pick up on those links.
However, contacts uploaded to Facebook are hashed for security and not visible to other users. So while your imported address book may influence suggestions, it does not expose your private data.
Shadow Profiles
There has also been speculation that Facebook maintains “shadow profiles” of non-users, based on data others upload about them. So if your friends all tag “John Doe” in photos and posts, he may be suggested once signing up for an account.
However, Facebook claims it does not create profiles for non-users. Any data about people not on Facebook is discarded within 90 days, as per company policy.
Other Possible Factors
Facebook has not confirmed exactly how much weight each of the above factors carries in People You May Know suggestions. It is likely a holistic analysis of different signals coming together.
Some other elements that may contribute include:
- Tagging in photos
- Being @mentioned in statuses and comments
- Sponsored event recommendations
- Matching advertiser targeting
- Other social graphs like Messenger and Instagram
- Linked devices and technologies
- Facial recognition on photos
Facebook has access to vast amounts of social data about users. Any connections or commonalities between you and others could potentially be mined for People You May Know suggestions.
Contacts You May Not Know
Interestingly, People You May Know does not only recommend existing offline connections. It may also suggest new contacts you have things in common with but have never met.
For example, if you move to a new city, it may recommend local residents with similar interests and backgrounds as a way to expand your network.
The “Know” in People You May Know does not necessarily imply you are already acquainted. It is about algorithmically determining people you might want to know based on shared attributes.
Maintaining Privacy
If certain People You May Know recommendations seem too personal or intrusive, you can remove the connection signal by:
- Unfriending or disconnecting from mutual friends
- Leaving shared Pages or Groups
- Unlike Pages that imply preferences you’d rather keep private
- Tightening privacy settings around work and education info
- Keeping location services disabled
You can also proactively block someone to prevent them showing up as a suggested friend. Or delete contacts you do not want informing Facebook’s suggestions.
Conclusion
Facebook’s People You May Know feature uses sophisticated algorithms to recommend new connections. It analyzes mutual friends, work/school affiliations, location, likes, events, contacts, and other signals to determine relevance.
Having more common touchpoints with another user makes a recommendation more likely. But even a few key data points overlapping can trigger a suggestion.
While useful for networking and meeting new contacts, People You May Know can also feel intrusive or surface unwanted connections. Maintaining privacy settings and curating visible profile data can help users control what information informs the recommendations.