Granovetter’s weak ties and the people you haven’t met yet.

Mark Granovetter published “The Strength of Weak Ties” in the American Journal of Sociology in 1973. He was writing about how people find jobs — specifically, about the mechanism by which job information travels through social networks. The paper has been cited more than 60,000 times. It is one of the most cited papers in sociology. Its core finding is about the structure of social networks, and the structural insight it produces applies with equal force to how people find romantic partners. Almost nobody has made that application explicitly.

This is a piece about the Granovetter finding, what it actually says, and why it matters for understanding the sociology of meeting people.

The core finding

Granovetter’s finding emerged from his 1969 doctoral research on professional job changers in a Boston suburb. He asked his subjects how they had found their current position. The majority had found it through personal contacts rather than through formal channels (newspaper ads, employment agencies). Among those who used personal contacts, most described those contacts as acquaintances — people they saw occasionally, perhaps a few times a year — rather than close friends.

This was counterintuitive. People looking for jobs might be expected to mobilize their closest relationships first, people who know them best and are most motivated to help. But close friends, Granovetter’s analysis showed, were less useful than acquaintances for accessing job information. The explanation became the theory.

Strong ties — close friends, family members, people you see regularly — cluster. They know who you know. The social network of a close friend overlaps substantially with your own. The information available in that network, including job openings and professional opportunities, is largely information you already have access to. Close ties are redundant as information sources.

Weak ties — acquaintances, people you see infrequently, people who move in different circles — bridge clusters. They have connections into social networks you don’t access. The information available through them is genuinely novel — it comes from outside your current network. This is why job seekers systematically found jobs through acquaintances rather than close friends. The acquaintances had access to information the job seeker’s close network didn’t.

Applied to partner finding

The application to partner finding is structural, not metaphorical. How do couples meet? The answer varies historically and culturally, but in contemporary Western contexts, the major channels include: apps and online platforms, mutual friends, bars and social venues, work, school and university, and repeated-context encounters (gym, neighborhood, recurring social settings).

The “mutual friends” channel is worth examining through the Granovetter lens. When couples describe meeting through mutual friends, the mechanism often involves a second-degree connection: Person A knows Person B, who knows Person C. A is introduced to C through B. The relevant question is whether B is a strong tie or a weak tie of A.

Strong ties — close friends — know your social network well. The people they might introduce you to are often people already in your network, or people in adjacent clusters you would encounter anyway. Weak ties, by contrast, bridge to people outside your current network. The acquaintance you know from a former job, the friend from a professional event, the person you know distantly from a community organization — these people have access to social clusters that don’t overlap with your primary network. Their introductions produce genuinely novel connections.

Cacioppo et al.’s (2013) data on how couples meet shows that meeting through mutual friends remains a significant pathway, but the category is broad enough to encompass both strong-tie and weak-tie introductions. The Granovetter logic suggests the weak-tie introductions are more likely to produce novel matches — people you wouldn’t have encountered within your existing network.

The network investment argument

The practical implication that follows from Granovetter’s analysis is not that you should attend more networking events in the transactional sense of the term. The implication is structural: maintaining diverse weak-tie connections provides access to social clusters your strong-tie network doesn’t reach.

Most people, when they think about investing in social relationships, think about deepening existing ones — spending more time with close friends, building stronger bonds with existing contacts. This is valuable for friendship and wellbeing. It is not, according to Granovetter’s analysis, where novel connections come from.

Novel connections — introductions to people outside your current network — come from the maintenance of bridge connections into other clusters. This requires a different kind of social investment: staying in light contact with a broad set of acquaintances across different contexts (former colleagues, people from earlier life chapters, contacts from different professional or recreational communities). The intensity of these relationships is low. The network function they serve is high.

The counterintuitive implication: the people most likely to introduce you to your next significant partner are probably not your closest friends. They are the people you see occasionally, who move in circles adjacent to but not identical with your own.

Robert Greene’s parallel observation

Greene’s work — The 48 Laws of Power and The Art of Seduction in particular — identifies the strategic value of broad social investment and the maintenance of diverse connections across different social contexts. Law 18 of 48 Laws: “Do Not Build Fortresses to Protect Yourself — Isolation is Dangerous.” The accompanying analysis in Greene’s text is about the value of remaining embedded in broad social networks rather than retreating into narrow circles.

This observation converges with Granovetter’s structural analysis despite coming from a completely different tradition. Greene is a practitioner-strategist drawing on historical observation; Granovetter is a sociologist working from empirical data. That they reach compatible conclusions through different methods is worth noting, though it doesn’t validate Greene as an empirical researcher. He isn’t one. His framework is not the evidence base here.

What Greene’s work does is provide a practical articulation of the network diversity principle in terms that have reached millions of readers. The 48 Laws has sold over five million copies. The observation about isolation being dangerous is not derived from the academic network literature, but it is consistent with it.

Marc Andreessen and network effects

Andreessen’s technology framework — that the value of a network grows nonlinearly with the number of connected nodes, following Metcalfe’s law — has an informal parallel in personal social networks. The person who maintains connections across twenty different social clusters doesn’t simply have more potential contacts than the person who maintains connections across five clusters. They have more bridge connections, more paths to novel information, and more potential introduction pathways to people entirely outside their current social world.

This is the network analysis of what Granovetter described empirically. The social analogue of Metcalfe’s law is not about the quantity of connections but about their diversity and their bridge function. A network of 500 strong ties within a single cluster is less structurally valuable, for the purpose of accessing novel people and information, than a network of 100 weak ties distributed across ten different clusters.

This is a description of network structure, not a prescription for relationship priorities. Close friendships are not less important because they are redundant as information bridges. They serve different functions — emotional support, shared history, mutual investment — that are not captured by network analysis. The point is not that weak ties should replace strong ties, but that weak ties serve a function that strong ties cannot, and that most people underinvest in them.

The actionable structure

The Granovetter research does not say “go to more events.” It describes the structural conditions under which novel connections become likely. Those conditions include: maintaining contact with a diverse set of acquaintances across different social and professional contexts, staying embedded in multiple communities rather than concentrating social investment in a single cluster, and remaining open to introductions from weak-tie contacts.

The platform-era complication: dating apps have partially replaced the weak-tie introduction mechanism by providing direct access to people entirely outside one’s social network. Apps are, in network terms, a radical bridging mechanism — they connect users to people with zero social overlap. Granovetter’s analysis still applies to non-app meeting channels, which remain significant, but the app ecosystem has created a partial substitute for the weak-tie introduction pathway.

The partial nature of that substitution matters. Cacioppo et al. (2013) found that relationships that begin online show different relationship quality trajectories than those formed through offline channels, including mutual-friend introductions. The algorithm-mediated match is not the same as the socially embedded introduction. The social context in which a weak-tie introduction occurs — shared acquaintances, shared community, mutual social accountability — shapes the subsequent relationship in ways the app match does not.