There’s someone you like.
You met them a year or two ago, at work or at a party or through the person you were dating at the time. The conversation was good. You both said the thing people say, which is that you should really do this more often.
You have seen them maybe six times since. You like them exactly as much as you did. And they are still, by any honest accounting, not your friend.
Nothing went wrong, which is the confusing part. There was no falling out, no bad first impression, no moment where either of you quietly decided against the other. The relationship just sat where it was and stayed there.
In 2018, a communication researcher at the University of Kansas published a paper that explains what happened, and the explanation has almost nothing to do with either of you.
Someone finally counted the hours
Jeffrey Hall wanted a number.
Everyone agrees friendship takes time. Nobody had actually measured how much. So Hall built two studies designed to put a figure on it.
The first surveyed 355 adults who had moved somewhere new within the previous six months. Each picked one person they had met since arriving, estimated the hours they had spent together, and placed the relationship somewhere on a scale running from acquaintance to close friend.
The second followed 112 first-year students across their first nine weeks on campus, checking in three separate times about two people each of them had just met.
Combining the two, Hall estimated it takes somewhere between 40 and 60 hours together to turn an acquaintance into a casual friend, another 80 to 100 to get from there to a friend, and more than 200 hours to become close.
Two hundred hours. That is five full working weeks, spent entirely on one person, on top of everything else your life already contains.
Hall’s own summary of it was blunt. “You can’t snap your fingers and make a friend,” he told his university’s news service.
The hours that count are the ones nobody schedules
Here is where the number gets worse instead of better.
Not all hours qualify. Hall found that time spent working alongside someone contributes far less than time spent doing nothing in particular together. The hours that move a relationship are the ones spent joking around, catching up, and talking about things that have no agenda attached.
Which explains a pattern almost everybody has lived through. Some participants reported spending hundreds of hours with colleagues and still filed those people under acquaintance.
The clock was running the whole time. It just wasn’t counting.
There’s a floor, too. Relationships that never got past acquaintance had usually logged no more than about 30 hours total, roughly what you would accumulate sitting in the same room a few hours a week and never once seeing each other outside it.
Hall described the design plainly when the study came out. “I was looking for cut-off points,” he told Psychology Today, meaning the thresholds where the odds of moving up a level crossed fifty percent.
So the real requirement is 200 hours of a specific and increasingly rare substance: unstructured, unproductive, unbillable time in which two people are doing nothing a calendar would recognize as an event.
More Bolde Stories
Two hundred hours is only half the requirement and the calendar is the other half
The part that gets dropped from every viral version of this study is that the hours have to be compressed.
Hall did not just count hours. He counted them against time. In his data, the odds of two people crossing into good or best friendship passed fifty percent at roughly 119 hours accumulated over three weeks, or about 219 hours accumulated over three months.
Run that second figure through a calendar and it comes out to something close to 73 hours a month with one person. For three consecutive months.
Among the first-year students, the ones who became close roughly doubled their time together at every check-in. Some of them were spending a third of their waking hours in a month with a single friend.
That’s not a schedule. That’s a dormitory.
And it is why the friendships almost everyone describes as their deepest were formed before the age of 22. School is an institution that manufactures concentrated unstructured hours and hands them out for free. Then one summer it stops, and nothing in adult life is built to replace it.
The average American now has about ten hours a month of friend time to spread across everyone they know
This is where the number turn into a diagnosis.
Researchers Viji Diane Kannan and Peter Veazie went through eighteen years of American Time Use Survey data to track how much time people actually spend with other people.
What they found for friends specifically: an annual daily average of 60 minutes in 2003, falling to 20 minutes by 2020. Hours worked per week showed up in their analysis as a structural constraint on social connection, which is a very restrained way of saying that the job takes the hours first.
Twenty minutes a day is about ten hours a month, and that is the entire budget, divided among everyone you would call a friend.
Now put the two studies next to each other. If you took every single one of those hours and spent them on one person, and never saw anybody else, you would reach 200 hours in about twenty months. In 2003 the same budget would have gotten you there in under seven.
And Hall’s concentrated version, the 73 hours a month that actually produced close friendships in his data, is more than seven times the entire monthly allowance the average American now has.
These are not identical measures and the comparison is rough. Hall counted hours with one new acquaintance in an unusually social period of life. The time-use figures count in-person friend time across a whole population. But even generously interpreted, the gap runs to an order of magnitude.
Which is worth separating from the smaller, cheaper interactions that fill most of a day. Those still do real work, and the research on chatting with baristas and neighbors is genuinely encouraging. They are a different system entirely, running on a much cheaper budget.
Personality explains who is lonelier than their neighbor, not why the whole neighborhood got lonelier
The instinct, when adult friendship stalls, is to look inward. You have gotten worse at this. You are too closed off, too awkward, too introverted, too much.
Personality does matter, and it would be dishonest to pretend otherwise.
A preregistered meta-analysis by Susanne Buecker and colleagues pooled 113 studies covering more than 93,000 people and found that extraversion and neuroticism were both meaningfully related to loneliness, with correlations sitting a little under and a little over .35 respectively.
Those are real effects. They are also cross-sectional. They capture a ranking at one moment in time, telling you who among a group reports more loneliness than whom, which is a separate question from why the whole group shifted upward.
And that separate question is the one the headline is about. A correlation around .37 accounts for something like an eighth of the variation between people, which is not enough to carry a national trend on its own.
Personality traits move slowly at the population level, in small increments across generations. Friend time fell by two thirds in seventeen years.
So if you are asking why one person in an office feels lonelier than the person at the next desk, personality is a reasonable place to look. If you are asking why both of them feel lonelier than they did fifteen years ago, you have to look at what happened to their week.
Hall has made a version of this point himself. “There’s a strong negative correlation between hours worked and face-to-face conversation,” he said in a 2025 interview with the app TouchBase, noting that the pattern holds across two decades of time-use data and shows up internationally.
More Bolde Stories
The number is rougher than it sounds, and it is still the most useful one anyone has produced
It is worth being honest about what this study is.
It is two samples, 355 online participants and 112 college freshmen, from one research group. Most of the hour estimates were retrospective and self-reported, which is a generous way of saying people guessed. The thresholds are averages with wide variation around them. Two hundred hours marks a zone where the odds tip, and plenty of real friendships land on either side of it.
The first-year students in particular were living through the single most socially concentrated period a person is likely to experience, which makes them a strange baseline for the 34-year-old with two kids and a commute.
Hall himself has never claimed the number is a formula. He has consistently framed it as an argument for prioritization.
But the shape of the finding survives every one of those caveats, and the shape is the useful part. Closeness is a function of accumulated unstructured time. The accumulation has to be fast enough to build momentum. Work hours barely count. And whatever is happening to the national supply of unstructured time, it is heading in one direction.
What that reframes is the self-blame, which is what most people carry around instead of the numbers. The friendship that never quite happened with the person you liked failed a resource test that almost every adult relationship in the country is currently failing.
There is one variable in the equation you actually control, and it’s not the total. It’s the density. The same 200 hours spread thinly over four years does something different than the same 200 hours packed into a season, and Hall’s own advice to newcomers is to take the invitations early, because the window in which people are open to a new person closes within a few months.
Almost nobody has 200 hours sitting available, and treating the full number as a target is a reliable way to never start.
So ask a smaller one. Whether the next three months contain more hours with one specific person than the last three did.
That question you can answer by Friday.
