lightbulb The big idea
A referral doubles as a search filter. In LinkedIn Recruiter, a 1st-degree connection to an employee at the hiring company triggers an officially named spotlight — "Have company connections" — that a recruiter can switch on to isolate candidates with an inside tie2. LinkedIn has not disclosed whether the tie also feeds the underlying ranking model, so treat it as a filter that exists, not a hidden ranking boost.
What actually propagates
Two paths turn a referral from "someone I know" into "on the recruiter's screen."
The hand-off path. Your contact tells the recruiter or hiring manager to look at you. LinkedIn doesn't see or measure that conversation. It still happens off-platform for most referrals.
The visibility path. Even without a name-drop, the tie shows up as a Recruiter spotlight (see the callout above)2. One shared 1st-degree connection is enough to land you on that filtered list.
Both paths need the same thing: a real 1st-degree connection to someone inside the company. The rest of this card explains which connections matter most.
The 2022 experiment most guides missed
Between 2015 and 2019, LinkedIn's own data-science team, working with outside academics, ran a randomized experiment on the algorithm that recommends new connections. The scale is not typical for a social-science paper: roughly 20 million users, 2 billion new connections, and 70 million job applications in the treatment window1.
Then came the result, and it contradicted the folk advice to "network your closest contacts." The strongest causal effect on job mobility came from moderately weak ties: connections with around ten mutuals in common. Very-strong ties already know what you know, so they have nothing new to tell you. Very-weak ties can't vouch for you. The middle ground moved people into new jobs, and the effect was strongest in digital and higher-tech industries1.
The paper ran in Science in 2022, authored by Rajkumar, Saint-Jacques, Bojinov, Brynjolfsson, and Aral. LinkedIn ran the experiment on its own users, on a scale unusual for social-science research.
What the hire funnel actually looks like
The numbers in the diagram above come from one large U.S. corporation, tracked at administrative-record granularity in a peer-reviewed labor-economics study3. Referred applicants were 6% of the pool and 29% of the hires. That is a fundamentally different funnel — referrals surface a different pool, not the same pool treated better.
Retention follows the same pattern. Burks, Cowgill, Hoffman, and Housman studied nine large firms in 2015, spanning call centers, long-haul trucking, and high-tech/IT. Referred workers were 10–30% less likely to quit, and referred high-tech workers were about 20% more likely to develop a patent4. The authors ran four separate tests to rule out mentoring by the referrer as the cause. The advantage is selection: referrals surface candidates who fit the specific job better.
The 1973 finding still holds
Mark Granovetter's 1973 study surveyed Boston-area professionals who had recently changed jobs. Among those who landed a job through a personal contact, only 16.7% saw that contact "often"; 55.6% saw them "occasionally"; 27.8% saw them "rarely"5. Most job-finding contacts were weak ties.
The 2022 LinkedIn experiment replicated that pattern with causal identification, and pinned down the specific tie strength that matters most. Fifty years apart, the two findings agree.
Two things to act on today
Two places to start:
- Reach out to your moderately weak ties, not your closest contacts. Old colleagues at a prior employer, alumni from your program, former teammates who moved to your target company. Chapter 7 gives the outreach templates and the specific criterion for identifying which contacts to prioritize.
Read Chapter 7 arrow_forward - Score your profile on the fields recruiters' filters can detect. The Scorecard checks Introducible: whether a "Have company connections" filter at the companies you care about would surface you at all. Four minutes, no signup.
Take the Scorecard arrow_forward
Notes
-
Karthik Rajkumar, Guillaume Saint-Jacques, Iavor Bojinov, Erik Brynjolfsson, Sinan Aral. "A Causal Test of the Strength of Weak Ties," Science 377, no. 6612 (2022): 1304–1310. Five-year randomized experiment (2015–2019) run on LinkedIn's connection-recommendation algorithm covering ~20 million users, ~2 billion new connections, and 70 million job applications. Moderately weak ties (~10 mutual connections) were causally the most effective at generating job mobility; the effect was strongest in digital/tech industries. science.org/doi/10.1126/science.abl4476 ↑ ↑ ↑
-
LinkedIn Recruiter Help — "Spotlights in Recruiter and Jobs." Officially names Have company connections as a filterable spotlight: 1st-degree connections to employees at the hiring organization. LinkedIn does not disclose whether the same tie also feeds the underlying ranking model — treat "recruiters can isolate your company connections" as confirmed; "this connection boosts your rank automatically" as not confirmed. linkedin.com/help/recruiter/answer/a414283 ↑ ↑ ↑
-
Meta Brown, Elizabeth Setren, Giorgio Topa. "Do Informal Referrals Lead to Better Matches? Evidence from a Firm's Employee Referral System." Federal Reserve Bank of New York Staff Report No. 568, 2013 (peer-reviewed version: Journal of Labor Economics 34, no. 1, 2016: 161–209). Single large U.S. corporation, administrative panel data. Referred applicants were 6% of the applicant pool, 21% of interviewees, 27% of offer recipients, and 29% of hires. Referred applicants were 7.3 percentage points more likely to be interviewed than job-board applicants and 2.4 points more likely to receive an offer, all significant at the 1% level. newyorkfed.org/…/sr568.pdf ↑
-
Stephen V. Burks, Bo Cowgill, Mitchell Hoffman, Michael Housman. "The Value of Hiring through Employee Referrals," Quarterly Journal of Economics 130, no. 2 (2015): 805–839. Nine large firms across call centers, long-haul trucking, and high-tech/IT; hundreds of thousands of workers and millions of applicants. Referred workers were 10–30% less likely to quit; referred high-tech workers ~20% more likely to develop a patent; profit per referred worker ~25% higher. Four separate tests ruled out "treatment" (mentoring from the referrer) — the advantage is selection. doi.org/10.1093/qje/qjv010 ↑
-
Mark Granovetter. "The Strength of Weak Ties," American Journal of Sociology 78, no. 6 (May 1973): 1360–1380. Of professional/technical/managerial workers in a Boston suburb who found a job through a personal contact, 16.7% saw that contact "often" (at least twice a week), 55.6% saw them "occasionally" (more than once a year, less than twice a week), and 27.8% saw them "rarely" (once a year or less). Foundational sociology paper; ~50,000+ citations. jstor.org/stable/2776392 ↑
Five primary-source citations, all Tier 1: peer-reviewed research (the 2022 Science weak-ties experiment, the QJE 2015 referral-hire study, the NY Fed SR568 replication, Granovetter's foundational 1973 AJS paper) plus LinkedIn's own official Recruiter Spotlights documentation. What this card deliberately does not claim to know: whether "Have company connections" also boosts your rank inside LinkedIn's ranking model, an exact modern (post-2020) "referrals are hired X× more often than cold applicants" ratio, and any specific median 2nd-degree network size — all flagged in the underlying research file.