This is for the PhD student who realised last Tuesday that they hadn't had a non-transactional conversation in nine days. It's for the postdoc who moved cities for a lab and now eats lunch at their desk. Here's how to make research friends without pretending you're doing something else.
Why 'networking' advice doesn't help you make actual friends
Most advice about "building your network" is aimed at your CV, not your week.
It tells you to introduce yourself at conferences, follow up on LinkedIn, and keep a spreadsheet of contacts. None of that produces the person you text at 10pm when your model won't converge. Networking optimises for weak ties. Friendship needs something else.
The other problem is that networking advice treats every interaction as a one-shot. You meet someone at a poster session, exchange emails, and then… nothing. You both go home. There's no next Tuesday.
If you want research friends — actual friends, not contacts — you have to stop borrowing frameworks from sales.
Friendship needs repetition, so design for recurring contact
The single biggest predictor of a friendship forming is repeated, unplanned contact with the same person over time. Sociologists have been saying this for decades, and anyone who made their closest friends in undergrad already knows it: proximity plus repetition plus low stakes.
Grad school breaks all three. Your cohort shrinks. Your labmates are on different projects. You start working from home two days a week. The scaffolding that used to produce friendship by accident isn't there anymore.
So you have to design for it. That means picking one or two recurring things — a weekly writing session, a monthly lunch, a standing coffee — and showing up even when you don't feel like it. Especially when you don't feel like it.
We built Research Lunch Club around a monthly cadence for exactly this reason. Once is a meeting. Every month is a friendship-in-progress. You can see how the lunches are matched if you want the mechanics.
Shared-context shortcuts: peers who get the research life
The other thing that makes research friendships different is context.
You can be close to people outside academia and still spend a huge amount of energy translating. Explaining what a preprint is. Explaining why your advisor's silence is scary. Explaining why you're happy and miserable about the same paper. It's not that non-researchers can't understand — it's that the translation tax is real, and it adds up.
A research friend skips the tax. When you say "reviewer 2" they laugh in the right place. When you say "I'm on the market" they know what that costs. When you say "my PI hasn't replied in three weeks" they don't say "just send a follow-up!"
This is why peer groups across disciplines often work better than you'd expect. A computational biologist and a UX researcher don't share methods, but they share the shape of the week. Deadlines, revisions, funding anxiety, the particular loneliness of a long project.
If you're looking for that kind of peer, cross-disciplinary formats are a shortcut. Team-science communities are another; the upcoming SciTS conference draws people who explicitly care about how researchers work together — check the site for 2026 dates.
Where to start this month
Here's the smallest useful version of this advice.
Pick one recurring thing in your city, and commit to three iterations of it before you decide whether it's working. Three is the minimum for repetition to do its work. One meeting tells you nothing. Three starts to tell you whether these are your people.
If your department runs a writing group, join it. If there's a Slack for postdocs in your field, post in it, don't just lurk. If a friend-of-a-friend runs a monthly dinner, go. And if none of those exist where you are, you're the one who has to start something — or borrow a format that already works.
RLC currently runs in cities across North America, Europe, and Asia-Pacific. You can browse the cities we run in to see if yours is on the list, and the FAQ covers the common questions about how it works — group size, matching, cost, all of it.
What to expect from a research friendship, honestly
One last thing, because it saves disappointment.
Research friendships often look quieter than the friendships you had at eighteen. They're not built on constant contact. They're built on the fact that when you email someone eight months later saying "I'm thinking about leaving my postdoc, can we talk?" they say yes, and mean it.
Many postdocs find that two or three of these people carry them through the hardest years. Not a network of forty. Two or three. So the goal isn't volume. The goal is finding the handful of people who get it, and then keeping the door open.
That starts with showing up somewhere, more than once, with people who share the shape of your week. If you're a researcher looking for that group, the next round of lunches is open for sign-ups — you can apply to join a lunch group. Pick a month, show up three times, and see who's still at the table.