Friendship takes time to build, but the numbers are less tidy than social media makes them look. Jeffrey A. Hall’s 2019 study in the Journal of Social and Personal Relationships found that time spent together predicted increasing friendship closeness, with different estimates depending on the sample and the level of friendship being measured.
In the longitudinal student sample, acquaintances were more likely than not to become casual friends after about 43 hours together, while the transition from friends to good or best friends passed the 50% mark at 119 hours. Hall’s broader conclusion was that, although close friendship can sometimes develop earlier, 200 or more hours was a plausible estimate for friends becoming good or best friends over roughly six weeks.
That figure should not be treated as a universal clock. Hall studied different populations, including first-year university students and adults who had recently relocated, and his own paper notes that friendship development depends on more than elapsed hours.
The widely shared 11-3-6 shorthand comes from a different source entirely. A 2022 Fisherman’s Friend-commissioned survey of 2,001 UK adults, analyzed by Robin Dunbar, reported an average of 11 interactions, each lasting a little over three hours, with friendship taking about 5.5 months to become established.
It is an interesting survey result, not a peer-reviewed law of friendship. It also measures something different from Hall’s estimate for moving from friendship into the good-or-best-friend category, which is why the two sets of numbers should not be treated as interchangeable.
The more revealing part is what those hours are actually building.
A certain kind of midlife adult can have a phone that pings throughout the day, dinners booked weeks ahead and a genuine reputation for being good company, yet still struggle to name the person they would call when something goes badly wrong at two in the morning.
That does not automatically mean the person is lonely, unhappy or secretly dissatisfied with every friendship they have. It means the width of a social network tells only part of the story about what happens inside it.
The dominant cultural story says that people with plenty of acquaintances and very few confidants must have chosen breadth over depth. They networked instead of connecting. They collected people instead of letting anybody close.
That interpretation may fit some lives, but network size alone cannot establish it.
Robin Dunbar, currently Emeritus Professor of Evolutionary Psychology at the University of Oxford, has spent decades studying the limits and structure of human social relationships. His work distinguishes a broad network of meaningful relationships from a much smaller core of people with whom the strongest bonds are maintained.
In a 2022 Forbes interview about friendship and wellbeing, Dunbar described an overall network of roughly 150 meaningful relationships and an inner core averaging about five close friends. Five is an average rather than a required quota, but the distinction between outer network and inner support group matters.
A person can be unusually good at maintaining the outer network. They remember birthdays, organize dinners, answer messages, introduce people, solve practical problems and show up when somebody else needs something.
Those are real forms of friendship. They can also become the role a person is best known for playing.
This is where usefulness becomes interesting. Being the reliable friend gives people repeated contact, trust and a reason to stay connected, but those things do not automatically require the reliable person to reveal much of themselves.
Hall’s research points to a narrower version of the same distinction. Time together predicted friendship closeness, but the kind of interaction mattered too. Catching up, checking in, joking and meaningful conversation predicted changes in closeness in the student sample beyond hours alone, while simply spending time together in obligatory settings such as work or school was not consistently associated with greater closeness.
That does not prove that useful people are secretly avoiding intimacy. It does show why a person can accumulate enormous amounts of social contact without every relationship becoming equally close.
For many adults, midlife makes the distinction easier to miss. Work becomes demanding, children or other family responsibilities can consume evenings, parents may need more help, partnerships require time, and the unstructured hours in which friendships once deepened become harder to find.
The relationships that fit inside the existing timetable therefore have an advantage. Colleagues are already nearby. Neighbours share geography. Parents may repeatedly see the same people through their children’s activities. Longstanding friends already have years of history behind them.
None of those relationships is inherently shallow. The point is that proximity can keep a relationship alive without automatically changing the role a person plays inside it.
This is also where internet psychology can become too confident. A wide social circle with few confidants is not enough information to diagnose an attachment style, identify a childhood wound or conclude that somebody is afraid of vulnerability.
Sometimes the explanation is less dramatic. A person has spent years being the one who copes, organizes, listens and helps, and asking somebody else to take that role for an hour now feels unfamiliar.
The role has genuine rewards. Being trusted feels good. Being dependable becomes part of identity. Other people may sincerely value the competence and steadiness that person brings into a room.
The cost appears only when the traffic runs in one direction for too long.
The useful friend knows who is getting divorced, who has a difficult manager, whose parent is ill and whose child is struggling at school. The harder question is how many of those people know the equivalent details about the useful friend.
That is what being out of practice at being known can mean without turning it into a diagnosis. The capacity for closeness may still be there; the habit of offering somebody else the unresolved, inconvenient version of life may simply have had less practice.
Dunbar told Forbes that the number and quality of close friendships in the small inner core are strongly associated with psychological and physical wellbeing. That does not mean everybody needs exactly five friends, or that an acquaintance-rich life is somehow fake. It means the closest layer of a network appears to do a different job from the wider one.
So the 2am test is useful mainly as a thought experiment. It separates social availability from social depth.
The practical shift is not necessarily to meet more people. For somebody whose calendar is already full, that may reproduce the same pattern at a larger scale.
A more meaningful change can happen inside relationships that already exist: staying on the phone after the practical reason for calling has ended, admitting that something is unresolved before there is a solution, or allowing another person to help without immediately repaying the favour.
Those moments are small enough to look insignificant. Over time, they change what information a friendship contains.
That is the useful insight hiding underneath all the friendship formulas. Hours matter, but they are not interchangeable. A hundred hours of being impressive, entertaining or indispensable do not necessarily create the same relationship as a hundred hours in which two people gradually learn what the other sounds like when neither of them has the answer.
Being liked by many people and deeply known by very few is therefore not automatically a failure of character. For some adults, it is what a successful social role looks like after years without much pressure to change it.
Sometimes the people who could become the 2am calls are already somewhere in the wider circle. What is missing is not another dinner reservation, another contact or another perfectly handled favour, but enough unedited moments for somebody on the other end of the phone to know why the call came at two in the morning.