Verke Editorial

Making friends as an adult: why it's harder now, and a plan that actually works

Verke Editorial · September 14, 2026

You have colleagues. You have a partner, maybe, and a family, and a group chat from university that has gone quiet in the way group chats do — a birthday every few months, a photo of somebody's baby, a thread about a reunion that never gets planned. You moved, or they did. Everyone had kids, or you did. And somewhere in the last few years you noticed that if you had a genuinely bad Tuesday, you would not know who to call. Not because nobody would answer. Because you would not know who it was allowed to be.

The second thing you noticed is that wanting this is embarrassing. Wanting a partner is a normal adult project with apps and advice columns. Wanting a friend, at thirty-four, feels like admitting to something — as if everyone else got theirs assigned at twenty-two and you missed the day. So you don't say it, and nothing changes, and the group chat stays quiet.

This article takes the embarrassment off the table by explaining why adult friendship stopped happening on its own, and then gives you a plan — a ladder, a rule of three, a script, and a way of counting the answers — that treats it as what it mostly is: a logistics problem, with a small fear attached. If the fear turns out to be the bigger part, we say so near the end, and point you somewhere better.

Nobody made friends at school. School made them for you. Sociologists who study how friendships form keep coming back to the same three conditions, and it is worth seeing them side by side, because adult life quietly removes every one.

Your twenties ran on all three by default. Then the defaults expired. You moved to a city for a job; the job is remote two days a week; the people you see most are people you are paid to see. Nothing about you changed. The machine that used to produce friends was switched off, and nobody told you it was a machine.

There is also the matter of time, and here the research is humbling. Work by the communication researcher Jeffrey Hall on how long friendships take to form suggests something like tens of hours of time together before someone moves from acquaintance to casual friend, and a great deal more — well over a hundred — before they count as close. The exact figures depend on how you ask, but the shape is not in doubt: one good conversation is nothing. Friendship is a volume business, and adult life gives you the volume in thimbles.

Put those together and the mystery dissolves. You are not worse at this than you were at twenty. You are running the same program on a machine with the defaults turned off and the clock running at a tenth of the speed. The plan below turns the defaults back on, one at a time, on purpose.

Before the plan, the thing that stops most people from using any plan. It goes like this. Wanting friends feels like evidence that something is wrong with you. So you hide it — from other people, and slightly from yourself. Hiding it means never making the first move, because a first move would reveal the wanting. Nothing changes, which feels like more evidence. The loop tightens a notch every year.

Cognitive behavioral therapy would ask you to find the sentence that runs the loop, and for adult friendship it is almost always a version of this one: "If I have to try, it doesn't count."Real friends, the story goes, happen naturally. Engineered friends are a consolation prize, and everyone can tell.

Look at that sentence next to the section above and it falls apart. Your old friendships did not happen naturally. They happened because an institution put you in a room with the same forty people for a thousand hours. That was engineering; you just were not the engineer. Doing it deliberately now is not a downgrade. It is the same thing with the scaffolding visible.

Two more that are worth catching, because they sound like observations and are actually predictions: "everyone my age already has their people" (they mostly don't; they are mostly running the same loop as you, and would be quietly relieved if somebody made the first move), and "it would be weird to ask" (the ask is weird for about four seconds, and then it is a plan for Thursday). If the wanting is the embarrassing part, our article on why you can feel disconnected even around people explains what loneliness does to the part of your brain that reads faces — it makes you a worse judge of whether people like you, in a specific direction.

CBT has a tool for things that feel like one enormous step: you break them into rungs small enough that each one is only slightly uncomfortable, and you climb them in order. Therapists use this for phobias. It works for friendship too, because "make a friend" is not an action. It is the name for the top of a ladder, and the reason you have not started is that you have been trying to jump to it from the ground.

Nothing becomes a friendship on one contact. Not one coffee, not one great evening at somebody's wedding, not one conversation where you both said "we should do this again" and meant it. The single most common failure in adult friendship is a wonderful first meeting followed by nothing, because both people assumed the other would follow up, and both were waiting to see if it counted.

So the rule: before you make a first invitation, decide what the second and third will be. Not who else you might invite — the same person, and the rough shape of the next two contacts. If the coffee goes fine, the second touch is the class on Thursday, and the third is a walk in three weeks. You do not have to announce this. You just have to know it, so that when the coffee ends well you are not standing there wondering whether it was real.

The rule does something else, too: it takes the pressure off the first meeting. A first coffee that has to prove something is a job interview. A first coffee that is one of three planned touches is a coffee. People can feel the difference, and they relax into the second kind.

Most invitations fail before they are sent, because they are written to be un-rejectable, and an un-rejectable invitation is a vague one. "We should grab a coffee sometime" cannot be declined, which is why it cannot be accepted either. A good invitation has three parts, and the third is the one people leave out.

Now the part that changes people's minds. Before you send the first invitation, write down a number: the probability, in your honest estimate, that this person says no. Most people write something between sixty and ninety percent. Then send it, and record the actual answer. Do this for every invitation you make for a month — different people, different rungs — and keep the tally somewhere you will see it.

This is a behavioral experiment, the workhorse of CBT: you take a belief that feels like knowledge, turn it into a prediction with a number on it, and go and get the data. What people find, over and over, is that the prediction was wildly pessimistic. The yes rate for a small, specific, easy-to-decline invitation is high. And the noes, when you ask what happened, are almost never about you — they are about a deadline, a sick child, a week that got away from someone. Which is why the rule of three touches has a corollary: ask each person once more, a few weeks later, before a no becomes a conclusion.

The tally also does something the loop cannot survive. It turns "nobody wants to be my friend", which is a feeling, into "seven of nine said yes", which is a fact. Feelings can argue with reassurance forever. They have a much harder time with a column of numbers in your own handwriting.

Everything above depends on the bottom rung, so it deserves its own section. You need a place that gives you back the three conditions from the start of this article: the same people, weekly, with something to do besides talk. Run that test against any option and most of the usual advice fails it. A one-off event gives you strangers once. A bar gives you different strangers each time. An app for making friends gives you a first date's worth of pressure with none of the shared activity to hide behind.

What passes: a weekly class in something you are bad at (being bad at it together is the setting where people drop their guard). A run club, a climbing gym, a five-a-side team, a choir. A volunteering shift with the same crew. A language exchange with a fixed group. A board-game night that meets on the same evening. A neighborhood garden. The parents at the same school gate, the same swimming lesson, the same Saturday football pitch. For a lot of people over thirty-five, their children's schedules are the best friendship machine they will ever have access to, and they never think of it that way.

Pick one. Not three; one you can actually attend for eight weeks in a row, because rung one only works with attendance. The first four weeks will feel like nothing is happening. That is the volume problem from earlier — you are accumulating hours — and quitting in week three because "nobody talks to me" is leaving the oven before it has warmed up.

If you have just moved, you have one advantage and one danger. The advantage is that arriving is the most socially legitimate reason to be looking there is: "I'm new here" opens doors that would be closed to someone who has lived on the street for ten years. The danger is that it expires, and that the loneliness of the first months makes you worse at using it — the guard goes up, you read a neutral face as a cold one, and you stay in.

So front-load it. Pick the recurring place in your first two weeks, before you have settled into an evening routine that does not include one. Say the words "I just moved here" early and often; people who have moved themselves will volunteer things. Give yourself eight weeks of attendance before you judge anything, and expect to make the first invitation around week five. If it is a new country as well as a new city, add a place where the point is language — the exchange, the class — because being visibly bad at the language is the fastest guard-dropper available.

If you have small children, or a calendar that looks like one, the usual advice assumes free evenings you do not have. Friendship inside constraints looks different, and works fine. The walk-and-talk: a regular half hour with the buggy, or the dog, or just the two of you and a route. The standing thing: the same slot every fortnight, agreed once and never re-negotiated, which removes the part of scheduling that kills adult friendships (the eleven messages to find a date). The piggyback: doing something you already have to do — the school run, the Saturday shop, the gym — alongside someone else who has to do it too.

The friendship you make at forty-one in twenty-minute pieces around other obligations is not a lesser version of the one you made at twenty in a shared kitchen. It is the same thing, adapted to the machine it is running on.

Everything above assumes that what stops you is the machine being switched off: no place, no repetition, no plan. For a lot of people that is the whole story, and the plan is enough. For some it is not, and it is worth being honest about which one you are, because the fix is different.

Here is the test. Picture yourself sending the twenty-two-word invitation from above to a real person you have met at a real place. If the obstacle that shows up is calendar-shaped — when, where, how to fit it in — you are out of practice, and the ladder will get you there. If what shows up is your heart rate, a very specific picture of them reading it and thinking something about you, and a strong wish to have never had the idea, that is not rustiness. That is social anxiety, and it is extremely common and well understood, but it does not respond to being pushed through with willpower. It responds to the same CBT tools used differently — the fear itself becomes the thing you test, in smaller steps than the ladder above. Our article on social anxiety and what actually helps is where to start, and there is a set of social anxiety exercises you can practice on your own that begins well below the first rung here.

A second honest limit. If the loneliness is part of a low mood that has taken the color out of everything — not just friendship, but food, work, the things you used to like — then a friendship plan is treating one symptom of something bigger, and it will feel like pushing a car uphill. That is a different conversation, and if it has been going on for weeks it is one to have with a doctor or a therapist, not with a plan on a website.

And a third, gentler one: not every rung-five coffee turns into a friend, and some of the people at your recurring place will simply not be your people. That is not a failure of the method. It is the method working — you found out in six weeks what would have taken a year of hoping.

The plan above is the easy part to read and the hard part to do alone, because every rung has a moment where the loop from earlier gets a vote. Judith is Verke's cognitive behavioral coach, and this is her home ground: name the specific situation you are avoiding (not "making friends" — the coffee with the person from Thursday), find the thought that makes it feel bigger than it is, and pick one thing small enough to honestly try this week.

In practice that means drafting the invitation with her until it reads like you; writing down your predicted no-rate before you send it; coming back afterwards with what actually happened, because she remembers what you tried and what the tally says; and rehearsing the two-minute conversation out loud, by voice, when the words are the part that will not come. She is practical rather than motivational. Nobody needs to be told that friends are good. People need to know what to type on Thursday. More about how Judith works.