Why doing something beats "grabbing drinks": the case for activity-first friendship

Somewhere along the way, adult friendship converged on a single move. You meet someone promising (at work, at a party, through a friend) and the next step writes itself: "We should grab drinks sometime." Or coffee, or dinner. The venue varies; the format doesn't. Two or three people, a place to sit, ninety minutes, and nothing whatsoever to do except talk.

It feels like the obvious move because it's the available one. But run the honest tape of how those evenings actually go: pleasant, a little effortful, vaguely interview-shaped, and (this is the tell) almost never repeated. The drinks were fine. The follow-up text never quite gets sent. Six months later you're "the person I got drinks with once" to each other, which is a category nobody has ever promoted to friend.

This isn't a you problem. It's a format problem, and it's the full case this page exists to make: doing something together systematically outperforms sitting face-to-face, for reasons you can name, and the fix is choosing activities on purpose.

Why is doing an activity better than getting drinks?

Because an activity carries the parts of socializing that sit-and-talk formats dump entirely on you. When there's nothing to do but talk, conversation becomes the whole job, with all the performance pressure that implies. An activity changes the physics: it gives everyone a shared thing to attend to, makes silence comfortable instead of alarming, produces conversation as a byproduct rather than a requirement, generates memories with actual hooks in them, comes with a built-in next time, and filters for a shared interest before anyone says a word. The connection still happens through talking. The activity is just the machine that reliably produces the talk.

The rest of this page takes those one at a time, including the cases where the thesis honestly fails.

The interview problem: when conversation is the entire job

Sit two people who barely know each other across from one another with nothing but menus, and you've built an interview and called it a hangout. Every silence is a problem someone has to solve. Every question is a small performance: asked to fill space, answered while composing the next one. Both people leave slightly tired and privately unsure how it went, which is evaluation posture, not friendship posture.

Notice what the format demands: continuous, self-sustaining conversation between people who don't yet share any material to sustain it with. That's the hardest conversational assignment there is, and we hand it to brand-new acquaintances by default. A small group softens this (a lull just means someone else picks up the thread), but even a group of five, seated with nothing to do but talk, is still doing conversation as a job. The problem isn't the headcount. It's that talk is load-bearing.

What does doing something side by side actually change?

There's a name for this in friendship research. Geoffrey Greif, who interviewed nearly 400 men for Buddy System, distinguished face-to-face friendship (connection built through conversation and disclosure) from shoulder-to-shoulder friendship: connection built by doing things side by side. A large cross-cultural study in PLOS ONE found the same split in the data, with activity- and club-based bonds on one side and intense, talk-maintained pairs on the other. That research got framed as a men's finding, but the mechanics under it aren't gendered at all. They're just how attention works, for everyone.

Side by side, three things flip. First, you get a shared object of attention: the trail, the boulder problem, the board state, the sauce that's about to break. Nobody is the object of attention, so nobody is being evaluated. You're both facing the thing, not each other. Second, silence becomes comfortable. When hands are busy, a quiet stretch isn't a stalled conversation; it's just the activity happening. The single most punishing feature of the drinks format simply disappears. Third, conversation emerges instead of being extracted. The activity generates its own material continuously (commentary, problem-solving, the running joke about your terrible knots), and the real conversations arrive on their own schedule, at the trailhead stop, over the drying dishes, without anyone having to summon them from nothing.

The memory effect: why activities become stories

Here's a test you can run on your own life, no citation required. Think of a newer person you've mostly met for coffee or drinks, and try to recall specific occasions. Odds are they've blurred into one undifferentiated composite, a warm smear of catch-ups you couldn't put in order. Now think of anyone you've actually done things with. The memories come back pre-labeled: the time the kayak flipped. The escape room you failed with four minutes of taunting to spare. The hike where you took the wrong fork and turned a loop into a death march.

Friendships run on shared stories: they're the reference points, the callbacks, the "remember when" that makes a relationship feel like it has a past. Activities mint them at a rate sit-and-talk formats can't touch, because an activity has stakes, surprises, and things that can go charmingly wrong. Nothing goes wrong at a coffee. That's precisely the problem: nothing happens at all, and a friendship with no events in it has nothing to be made of.

The recurrence advantage: activities come with a built-in next time

This one is the quiet killer, because it operates after the hangout ends. Drinks are a one-off by construction: when the night's over, the format is over, and a next time exists only if somebody re-initiates from scratch. That's how "we should do this again!" (said sincerely, by both people) dies in the intention gap. Each future meetup needs a fresh act of social courage, and courage is exactly the resource new acquaintances have least of.

Activities recur natively. A league has a schedule. A class has a syllabus. A club has a standing night; a run group has a standing route. Nobody has to invite anybody. The next time already exists, and showing up requires no more courage than remembering it's Tuesday. This matters because friendship is, at bottom, an accumulation problem: it runs on shared hours, a lot of them, and mostly the chosen, leisure kind you only get by showing up again and again. A format you must re-initiate every time leaks those hours at every gap. A format that recurs on its own banks them automatically, with the same people, week after week.

And there's a bonus working before anyone even arrives: self-selection. An activity filters for a shared interest before a word is spoken. Everyone at the climbing night likes climbing; everyone at the board game meetup opted into board games. You start with a guaranteed topic, a guaranteed overlap, and a guaranteed reason to come back. "Drinks" filters for nothing except politeness and a free Thursday, which is why you can have a perfectly nice evening with someone you have perfectly nothing in common with, and never find out until the conversation runs dry.

Where the drinks-beats-activities thesis honestly fails

Two places, and pretending otherwise would make this a worse argument.

First: some people genuinely connect best face-to-face. The quiet-coffee, long-conversation style of friendship is real, legitimate, and for some temperaments it's home. If two hours of unhurried talk is where you come alive, that's not a format error to correct, and coffee remains the best tool ever invented for deepening a friendship that already has material to draw on. The case here is against drinks as the universal opening format, not against conversation.

Second: activities with zero talk-space fail in the opposite direction. A spin class at 110 decibels, a movie, a concert: side by side, sure, but there's no channel. You can attend weekly for a year and know nothing about anyone. Shared attention only builds friendship when conversation can ride along with it. The failure modes bracket the answer: all talk-space and no activity is an interview; all activity and no talk-space is parallel solitude. The sweet spot is talk-compatible activities (walking, hiking, climbing, board games, cooking, crafting, conversational-pace running), where the activity absorbs the pressure and the talking still fits.

What activities are best for making friends?

The ones that pass a three-question test. Would you half-enjoy it alone? Then showing up doesn't depend on the social payoff, and you'll still be there in week three when the friendships haven't formed yet. Does it recur natively? A club, league, class, or standing night means no one ever has to re-initiate. Can you talk during it? If not, you're building attendance, not friendship. Two of three is workable. Three of three (a walking group, a climbing night, a board game meetup, a cooking club, a volunteering shift, a rec league with a bench to sit on) is the format doing most of the work for you. The test is a fast screen you can run in your head; when you're ready to weigh your real options one against another, here's how to pick the right activity.

Run the test and a useful inversion appears: the question was never "what should we talk about?" It was "what should we do?" Because the doing produces the talking, and never the other way around.

None of this is something we bolted onto AroundMe. It's closer to the product's middle name. The reason activity-first friendship stays rarer than it should isn't that people don't want it; it's that the format doesn't assemble itself. Someone has to gather the 4-to-6 people who already share the interest, set a concrete plan at a real meeting spot, and put it on a rhythm that repeats. The problem was never you. It's that this infrastructure doesn't exist by default, so we build it: the activity as the engine of the meetup, not the excuse for it. We didn't invent side-by-side friendship. We just refused to ship one more way of saying "let's grab drinks sometime."

Picture the version that works: six near-strangers crowded around a stovetop that's a little too small for the group, one of them insisting there's no such thing as too much garlic while another quietly rescues the sauce, the real conversation surfacing sideways between tasks, the way it only does when keeping it going isn't anyone's job. None of them is wondering whether there'll be a next time; it's already on the calendar. Nobody grabbed drinks. Everybody's coming back.

Sources: Geoffrey L. Greif, Buddy System: Understanding Male Friendships (Oxford University Press, 2008); Tamas David-Barrett et al., "Women Favour Dyadic Relationships, but Men Prefer Clubs", PLOS ONE (2015); Jeffrey A. Hall, "How many hours does it take to make a friend?", Journal of Social and Personal Relationships (2018).