Networking

A professional network is not a list of people you know. It is the set of people who would take an action on your behalf without you having to make a case for it. That set is produced by two things: work they observed you do, and enough repeated low-stakes contact that recommending you costs them nothing. Nearly everything sold as networking optimizes for contact volume, which is the input with the weakest relationship to the output.

What actually creates a usable tie

The strong version of a tie is co-observed work. Someone who has seen you handle a deadline, a bad quarter, or a decision with incomplete information has information about you that no conversation transmits. They can vouch for you because they have evidence. This is why the most valuable parts of most people's networks are former colleagues, and why the tie survives years of no contact.

The weak version is repeated, low-cost, non-transactional contact. It is cheap to maintain and it accumulates. The mechanism is reputational: when your name comes up, the person has a recent, specific, positive thing to say instead of a vague one.

Mark Granovetter's "The Strength of Weak Ties" is the foundational result here, and it is routinely misquoted. The finding is not that weak ties matter more than close ones. It is that weak ties reach further. Your close circle mostly knows what you know and who you know; the useful, non-redundant information arrives through acquaintances because they are embedded in a different part of the graph. A network's value is a function of its structural reach, not its headcount.

The two things worth being

Be legible and be a router. Legible means people can state in one sentence what you do and what you are unusually good at. If they cannot, they will not think of you when the relevant opportunity crosses their desk, no matter how much they like you. Vagueness is the most common defect in otherwise strong networks.

A router is someone who connects other people without an expectation of return. This is not altruism; it is positioning. The person who reliably makes good introductions occupies a structural hole, sees flow early, and accumulates obligation across an entire graph rather than in one relationship. The cost is low and the compounding is real.

Where the common understanding is wrong

Networking events select against the people worth meeting. The attendees are, on average, those with slack and something to sell. The people whose help would matter most are busy doing the work that made them worth knowing. Rooms organized around the *activity* of networking are thin markets for exactly that reason.

The second error is treating contact as maintenance. A quarterly "just checking in" message with no content is a request for attention disguised as generosity, and it reads that way. Sending something specific and useful once a year outperforms it, because it carries information rather than asking for a receipt.

The third is the tally. People who track favors given and owed end up with a network of transactions, and transactional ties do not survive the moment you need something unreciprocated — which is the only moment a network has to be worth anything. Do the useful thing, do not send the invoice, and let the ledger stay implicit.

The most reliable networking strategy is still to do visible work with competent people and not disappear afterward.