Why Jobs Come from Acquaintances of Acquaintances, Not Best Friends — The Strength of Weak Ties

Introduction — One Question That Overturned Common Sense

Think of the person who helped you land a job or make a career move. Was it a best friend, or "that person you talk to once in a while"?

Around 1970, Mark Granovetter, then a graduate student at Harvard, put exactly this question to white-collar workers in Newton, near Boston, who had recently changed jobs. The commonsense prediction is clear: the people who care about me most and will work hardest to help me are my close friends, so the decisive help should have come from them.

The result was the opposite. When those who had found jobs through personal contacts were asked how often they saw the person who gave them the tip — only 16.7% said "often," while 55.6% said "occasionally" and 27.8% said "rarely." More than 80 percent of the decisive information had come through weak ties. The 1973 paper that forged this finding into theory, "The Strength of Weak Ties," went on to become one of the most cited papers in all of social science. Today we read this elegant theory through to the 20-million-person experiment that came half a century later.

Why Weak Ties — The Logic of Bridges

Granovetter's explanation lies not in emotion but in structure. The logic runs in two steps.

Step 1: strong ties clump together. If I am close to both A and B, then A and B are likely to come to know each other as well. Time spent together and mutual similarity make it so (triadic closure). As a result, strong ties form dense clumps (clusters) in which everyone knows everyone. Most friends of a best friend are already my friends.

Step 2: so bridges are almost always weak ties. Consider the only line connecting two different clumps — the tie Granovetter called a bridge. If that line were a strong tie, triadic closure would spawn other connections around it, and it would no longer be the "only" line. A bridge linking two worlds is therefore, structurally, bound to be a weak tie.

Now layer the flow of information on top, and the picture completes itself. Inside a clump of strong ties, everyone knows the same people, hears the same news, and sees the same job postings. However much a best friend wants to help me, most of what they know, I already know. The acquaintance I see only occasionally, by contrast, lives in the clump of a different company, a different industry, a different city. Their ordinary everyday information is a new opportunity for me. The strength of weak ties comes not from the depth of affection but from the non-redundancy of information.

This structure generalizes well beyond job hunting: the paths along which new ideas and trends spread, the routes by which innovation crosses departmental borders in organizations, and, in reverse, the way a community woven only of strong ties drifts into an echo chamber. In later work Granovetter pointed out that weak ties are also a key to explaining the isolation of impoverished communities. Into a neighborhood without bridges, opportunity does not cross either.

Half a Century of Testing — and the 2022 LinkedIn Experiment

The theory had a long-standing weakness. The original study was a retrospective survey of 282 people, and most of the supporting studies over the following decades were correlational. If "people with many weak ties get good jobs" is an observation, the causal arrow could point the other way, or a third variable could be at work.

Causal verification arrived in 2022. A team of MIT, LinkedIn, and Harvard researchers (Rajkumar and colleagues), publishing in Science, exploited A/B tests of the LinkedIn "People You May Know" recommendation algorithm. Over five years, the mix of strong versus weak ties recommended to 20 million users was varied at random, and the researchers tracked the 2 billion new connections that formed and 600,000 job changes that followed. It is, in effect, the largest randomized experiment ever run on the weak-tie theory.

The results sided with the theory — while adding interesting twists.

• Newly formed weak ties were significantly more likely than strong ties to lead to a job change. The effect was confirmed causally.
• The relationship, however, was not monotonic but an inverted U. The most effective ties were the "moderately weak" ones — pairs with at least a few mutual acquaintances — while ultra-weak, near-random connections lost their punch. Even a bridge, it seems, gets crossed only when it rests on a minimum of trust.
• The effect was especially large in fast-changing industries such as IT. Where the freshness of information matters most, bridges are worth most — a result consistent with the theory.

Practice — Weak-Tie Maintenance for Introverts

If the word "networking" makes you tired, let me say this first: practicing this theory is not about scattering business cards. The practice of someone who understands the structure is quiet and high-yield.

1. Start with dormant ties. A more powerful, undervalued asset than meeting new people is dormant ties — former colleagues, old classmates, people from previous jobs. In research by the organizational scholar Daniel Levin and his team, the advice executives obtained by reconnecting with dormant ties was rated as fresher than advice from current ties, and trustworthy at the same time. Those people have been living in different worlds in the meantime, so their information is new, while the old trust remains. Once a quarter, one how-are-you message to a former colleague who happens to cross your mind — that is bridge maintenance.

2. Put yourself regularly in activities outside your clump. Conferences, study groups, communities, and side projects are the habitats where weak ties arise naturally. The key is not frequency but diversity. Meeting someone from another industry once is structurally worth more than meeting a coworker one more time.

3. Become a bridge first. The fastest way to earn trust in a network is to introduce two people to each other. "A former colleague of mine does exactly this kind of work — shall I connect you?" Around the person who has served as a bridge, bridges gather.

4. Lower the friction of staying in touch. Weak ties break not because of conflict but because "there is no pretext for reaching out." Sending a short reaction to their news (a talk, a job change, a post), and reporting back the outcome when someone has helped you — pretexts are not manufactured, they are picked up. If awkwardness worries you, one follow-up question from the conversation skills post is enough.

Closing — The Map of Affection and the Map of Information Are Different

It needs to be said clearly that this theory does not belittle strong ties. Comfort, support, real help in hard times — what carries the weight of life is still strong ties (as discussed in the boundaries post, protecting those few people matters for exactly that reason). What Granovetter showed is that there is one more map. On the map of affection, the people close to us stand at the center; but on the map of information, distant acquaintances are the frontier.

So the practice for today is not a burden. That person who suddenly comes to mind — someone you worked with years ago, now living in a different world — send them a how-are-you message. It is the smallest of investments, underwritten by one of the most cited papers in sociology.