LinkedIn’s suggested connections algorithm is designed to recommend people you may know or want to connect with on LinkedIn. The algorithm works by analyzing your existing connections, profile information, activity on LinkedIn, and other data signals to generate suggestions that are personalized for you.
How does LinkedIn’s suggested connections algorithm work?
LinkedIn has not publicly revealed the full details of how their suggested connections algorithm works, but they have shared some insight into the general factors that influence the recommendations:
- Your existing connections – People you are both connected to are more likely to be suggested.
- Connections of your connections – 2nd degree connections are commonly suggested.
- Location – People located in the same city or region may be suggested.
- Industry and companies – People who work in your industry or at companies you are connected to are often suggested.
- Groups – Fellow group members are likely to be suggested.
- Education – Alumni from your schools and universities are frequently suggested.
- Shared interests – LinkedIn analyzes profiles to match people with common interests.
- Profile views – People who have viewed your profile are more likely to be suggested.
- Recent activity – Interacting with someone’s content may cause them to be suggested.
In addition to these factors, LinkedIn uses data science and machine learning techniques to continuously refine and improve the relevance of suggestions based on billions of data points and feedback from their users.
What data does LinkedIn use?
Here are some of the specific data points LinkedIn may use in their suggested connections algorithm:
- Profile information – Industry, occupation, skills, education, location, etc.
- Connections – Your 1st, 2nd, and 3rd degree connections.
- Follower and followee relationships.
- Group memberships.
- Job history and companies worked for.
- Schools attended.
- Content interactions – Likes, comments, shares, etc.
- Searches on LinkedIn.
- Profile views and visitor data.
- Shared connections and contacts imported from email and other sources.
- Shared interests, hobbies, volunteer work, and other profile details.
LinkedIn may also use data and signals from Microsoft services and partners to enhance their suggestions. However, they state that they do not use sensitive data like health information or credit scores.
How are suggested connections ranked and prioritized?
LinkedIn uses the various data signals and machine learning algorithms to assign a relevance score to each potential connection suggestion. The suggestions are then ranked and served to users in descending order of assigned relevance score.
Some key factors that may increase a suggestion’s relevance score include:
- Closeness of the relationship based on connections, shared experience, and profile similarities.
- Number of shared 1st and 2nd degree connections.
- Frequency of profile views and interactions.
- Being a recent connection of someone in your network.
- New members of groups you have joined.
- Alumni of your schools working at your company.
LinkedIn is also able to tweak the algorithm over time based on acceptance rates of suggestions as well as direct user feedback on the quality of recommendations received.
How often are new suggestions generated?
LinkedIn does not share specifics on how frequently they refresh the suggested connections list, but it appears to be dynamic and update regularly based on new activity and changes in your network and profile details.
Some users report seeing new suggestions appear daily or even multiple times per day. The rate likely also depends on how actively you engage with LinkedIn.
Can you refine the suggestions you see?
LinkedIn provides a few tools to help refine and improve the relevance of the suggestions you see:
- Filters – You can filter suggestions by relationship, location, industry, company, school, and groups.
- Remove suggestions – You can remove irrelevant suggestions from the list.
- Manage preferences – Adjust settings for how often you’re open to suggestions.
- Provide feedback – LinkedIn will sometimes prompt you to give feedback on suggestions.
However, many users report still seeing irrelevant suggestions even after providing feedback. The algorithm does not always seem responsive to user input.
Does LinkedIn customize suggestions based on user demographics?
LinkedIn says they avoid using sensitive member data like gender, ethnicity, or sexual orientation to determine suggested connections. However, some of the data signals they do use, like alumni connections and group memberships, could indirectly factor demographics into the mix.
For example, a female user may receive more suggestions for other women who graduated from her college and are in similar career networking groups. So demographics can play a role in recommendations even without directly personalizing by gender, race, etc.
Do they suggest people you’ve already rejected?
In some cases, LinkedIn will resurface and re-recommend connections you previously ignored or rejected. This seems to happen most frequently with 2nd degree connections who are highly connected to your network.
Even if you pass on someone initially, LinkedIn may determine there is still a strong relevance based on your shared connections and profile details. However, each additional rejection will decrease the likelihood of seeing that suggestion again.
Does LinkedIn avoid suggesting competitors?
LinkedIn’s suggested connections are primarily based on professional relationships and networks, not competitive factors. In many cases, current or potential competitors in your industry are likely to organically show up as suggestions due to your shared connections, groups, interests, etc.
However, directly connected contacts like coworkers or clients are less likely to recommend connecting with known competitors. So while you may see some, your closest contacts can indirectly deter suggestions of direct competitors.
Can you game the LinkedIn algorithm?
While you can take certain actions to increase the likelihood of getting suggestions you want, gaming LinkedIn’s system is difficult:
- Tailor your profile to highlight connections, experience, and interests aligned with your target suggestions.
- Proactively connect to and engage with related 2nd degree contacts.
- Join industry and alumni groups of people you want suggested.
- Connect with those who often make valuable introductions.
However, since the algorithm uses complex data science and evolving machine learning models, any “shortcuts” are temporary at best. LinkedIn’s system has been trained on billions of data points and is designed to provide relevant suggestions long-term.
Does LinkedIn connect you with inactive accounts?
LinkedIn’s algorithm focuses on suggesting active, engaged accounts rather than inactive ones. Signals like current employment, recently published content, and profile viewing activity are prioritized.
However, you may still encounter some suggestions that turn out to be inactive accounts. This can happen when a connection of yours changes roles or stops using LinkedIn. The algorithm may take some time to catch these changes in activity.
Can you retrieve previously suggested connections?
Once you dismiss or reject a suggestion, there is no built-in way to retrieve or revisit those previously suggested accounts within LinkedIn.
However, there are a few workaround options:
- Use LinkedIn search to find their name or company.
- Check your list of profile visitors.
- Look through your connections’ lists of connections.
- Use a LinkedIn connection tracking tool or browser extension.
You can also try engaging more with their content and network to get them suggested again.
Conclusion
LinkedIn’s suggested connections use sophisticated data science to facilitate valuable professional networking and relationship building on the platform. But the algorithm is imperfect and may require some refinement and filtering on the user’s end to get the most relevant recommendations.
Understanding the basics of how LinkedIn suggests connections can help you improve the connections you receive and expand your professional network more strategically.