The Best Way to Make Friends as an Adult: One Habit That Actually Works

If you want the best way to make friends as an adult compressed into a single line: pick a context you can return to weekly, return to it for at least three months, and be the person who proposes the next thing.

The reason this sounds underwhelming is that it isn't a hack. There's no clever opener, no charisma move, no app that does the work for you. Every part of it is unglamorous. You go to the same Tuesday yoga class. You see the same five people. You eventually say hi. You eventually invite one of them for coffee. Repeat.

What makes this the best answer rather than just a fine answer is that it cooperates with how friendship actually forms. Jeffrey Hall's 2018 work at the University of Kansas estimated that moving from acquaintance to casual friend takes roughly 50 hours of shared time, and over 200 to become close. Hours don't accumulate at one-off events. They accumulate in the contexts you return to. A wine class you take once gives you maybe three hours. A weekly climbing night sustained for a year gives you fifty or sixty before you've even tried.

The second half — being the one who initiates — exists because of the liking gap, which I'll get to in a later section. The short version: both of you almost always want to hang out more than either of you is signaling. Someone has to break the stall, and statistically nobody else is going to.

The rest of this page exists because almost every popular piece of advice on this topic points somewhere else, and I want to be specific about why that advice is worse than this one.

Half the fights about friendship advice come from people optimizing for different things and not realizing it. So before defending the answer, let me be explicit about which metric I'm using.

There are at least three reasonable definitions of best. First, fastest: how do I have someone to text within a week? Second, deepest: how do I build the kind of friendship I could call at 2am? Third, most durable: how do I end up with friendships that survive moves, jobs, marriages, and the next decade of my life?

Fastest is usually optimized by something high-intensity and short — a weekend retreat, a bachelorette trip, a week-long bootcamp. Those produce 'instant intimacy' that often evaporates two weeks later when nobody is in the same room anymore. Deepest is optimized by vulnerability and self-disclosure, but it's not a starting move — vulnerability without prior trust mostly makes strangers uncomfortable.

Most durable is the one almost every adult actually wants, even when they say they want fast. People I know who tried to optimize for speed ended up with a string of warm five-week friendships that ghosted. People who optimized for durability ended up, two years in, with a small number of real friends.

Repeated proximity in a low-stakes context plus initiating the follow-up wins on durability decisively, does fine on depth (because the hour count gets you there), and loses on speed. If you need a friend by Friday, this is not the right framework — call an existing weak tie. If you want a real social life in twelve months that's still intact in five years, this is the right framework. I'm picking durability because that's the version of the question almost everyone is secretly asking.

Most articles answering this question give one of three pieces of advice. All three are worse than useless. Let me take them in turn.

'Just put yourself out there.' This sounds harmless and is, in practice, the reason a lot of people burn out. 'Out there' tends to mean novelty — try a new event every week, say yes to everything, broaden your circle. The Hall hours problem destroys this strategy. If you go to ten different things once, you have ten three-hour relationships with rooms full of strangers. If you go to one thing ten times, you have a thirty-hour relationship with the same handful of people. The second one produces friends. The first one produces fatigue.

'Be more confident.' This is the worst kind of advice because it sounds true and is unactionable. You cannot will yourself into confidence at a Saturday meetup. Confidence is largely a function of familiarity — you feel confident in contexts you've been in before, with people you've seen before. So 'be more confident' is, sneakily, just a restatement of 'go back to the same place enough times that you stop feeling weird.' Skip the middle step. Just go back.

'You have to love yourself first.' I understand why people say this and I'm not going to be mean about it. But empirically, lots of people with shaky self-esteem have great friends, and lots of people in pristine therapeutic shape are friend-poor. Self-work and friendship-work happen in parallel. Friendship often does some of the self-work for you. Waiting until you feel internally ready will keep you waiting indefinitely.

The usual advice fails because it sounds like a posture you can adopt. The thing that actually works is a schedule item. Put a recurring event on your calendar. Go to it. That's the lever.

Three pieces of research underpin why the recommended answer works, and it's worth being concrete about them because most popular friendship writing waves at 'science says' without naming anything.

First, Jeffrey Hall, 2018. Hall surveyed hundreds of adults and used statistical modeling to estimate the hours required to move from stranger to acquaintance to casual friend to close friend. The headline numbers — about 50 hours to become a casual friend, around 90 to be a real friend, over 200 to be close — aren't a strict formula, but the order of magnitude is what most adults underestimate by tenfold. The implication is useful: there's no way to skip the time on the clock. You can only choose contexts that let it accumulate.

Second, Erica Boothby and colleagues, 2018. Across multiple studies, Boothby's team showed something they called the liking gap. After a first conversation, people consistently underestimate how much the other person liked them. Both sides walk away assuming the warmth was one-sided. Both sides therefore don't reach out. This is the psychological reason the second move stalls, and it's why being the one who initiates is so disproportionately powerful — you're breaking a stall that exists by default.

Third, Robin Dunbar's research on friendship layers. Humans appear to sustain roughly five close relationships, fifteen good friends, and a wider circle of about fifty friendly acquaintances, with the numbers tapering past that. The relevance here is that you're not trying to manufacture a teeming social calendar. You're trying to fill five to fifteen slots, slowly, over years. That's a tractable target. Knowing the ceiling stops you from feeling like you're failing because you haven't built a friend-group of thirty.

Hall says the time is unavoidable. Boothby says someone has to break the stall. Dunbar says the target is smaller than you think. Put together, they describe exactly the strategy this page recommends. A friend of mine — call him T — only really believed the liking gap when he texted someone from his pottery class on a whim and got back a reply that started with 'oh thank god, I've been wanting to ask you the same thing for weeks.' That sentence is the shape of almost every adult friendship that ever takes off.

Here is the actual sequence, in order, with nothing skipped. Read it like a recipe.

Step one: pick exactly one repeating context. One. Not three. A weekly class, a weekly run club, a recurring board game night, a monthly choir if monthly is the most you can sustain. Constraints: it has to be the same group of humans every time, not a rotating cast, and you have to be able to show up consistently for at least three months. Bars, festivals, and one-off mixers do not count.

Step two: commit to twelve weeks of attendance before evaluating anything. This is the part most people fail. You will go three times, feel like nobody noticed you, and want to quit. Don't. At three weeks you have nine hours in the room with these people. That's not enough to know anything.

Step three: in weeks four through eight, start having one slightly-personal conversation per session. Not a deep one. Just slightly past surface. 'How long have you been doing this?' becomes 'What got you into it?' becomes 'What else are you into?' becomes 'Where did you grow up?' Self-disclosure ladders up gently. Match whatever level the other person offers.

Step four: pick one person you'd genuinely like to know more, and within forty-eight hours of a good conversation with them, propose a specific, concrete, low-stakes follow-up. Not 'we should hang out sometime.' Something like: 'There's a coffee place near the gym I keep meaning to try — Saturday morning?' Specificity is the unlock. Vague invitations get vague non-responses.

Step five: if they say yes, propose a third thing at the end of the second thing. This is the move almost nobody makes and it matters disproportionately. 'This was great — want to make it semi-regular?' The transition from acquaintance to friend is mostly the transition from one-off plans to recurring ones.

Step six: repeat with one or two more people from the same context over the next six months. You're not trying to befriend the whole group. You're trying to convert two or three people from same-room people into actual friends. Six months in, you'll know if it's working. It almost always is, when people actually do the steps.

When this strategy doesn't work for someone, it's usually one of five things, and the fix is almost always smaller than the person thinks.

The first is dosage failure. They picked a repeating context but only went four or five times before stopping. The fix isn't a new strategy; it's going back to the same place six more weeks. The hour count is the lever, and they haven't pulled it.

The second is context choice. They picked a context where the same people don't actually come back — a giant gym, a huge meetup group, a Friday-night bar, a co-working space with rotating membership. The fix is choosing something with structural return: a class that runs for eight weeks with the same roster, a club with a stable core group, a league with set teams. Stability matters more than glamour.

The third is the initiation gap. They've gone consistently for three months, recognized faces, had good conversations, and then waited for someone else to propose hanging out outside the context. Nobody does. The fix is sending one specific 48-hour text to one person. That's it. The whole stall is one text wide.

The fourth is over-reading early signals. They had one slightly cool conversation with someone and decided that person doesn't like them. Boothby's liking gap says that conclusion is statistically wrong most of the time. The fix is going back next week as if it didn't happen, because for the other person, it almost certainly didn't register the way you remember it.

The fifth is the secret one: trying too hard, too early. Showing up to week three with a long emotional disclosure or an aggressive proposal to be best friends scares people. Calibrated patience reads as warmth; emotional intensity at week three reads as something else. Slow is faster here than fast.

None of these are personality problems. All of them are habit problems. Habits are easier to change than personality, which is why this approach is the best way to make friends rather than just one option among many.