The Myth That Meeting More People Fixes Loneliness

Open any app built for loneliness and it will do the same thing: introduce you to someone new. A match, a match, another match. The implicit theory is that loneliness is a supply problem. You don’t know enough people, so the fix is to know more people.

This is close to backwards.

The research on adult friendship keeps landing on the same finding, in different populations, using different methods: the number of people you know matters far less than whether you see the same ones regularly. A 2025 study of college students coming out of the pandemic found that it was the quality and regularity of contact, not the sheer number of connections, that predicted whether students felt supported and well. The American Enterprise Institute’s 2021 survey of American friendship found something that points the same direction from the other end: the collapse has not been in acquaintance count, which people can rack up easily online, but in the small number of people someone would call a close friend. Twelve percent of Americans report having no close friends at all, up from 3% in 1990. That is not a discovery problem. Most of those people have met plenty of new people since 1990. What they don’t have is anyone they’ve kept seeing.

Why the myth persists anyway

It persists because meeting someone new is the part that can be turned into a product. An app can match you with a stranger and a venue. It cannot make you go back to the same trivia night four weeks running, or force the second person to also show up. So the tools cluster around the one moment that’s engineerable — the introduction — and treat the rest as the user’s problem.

Robert Putnam’s Bowling Alone made the underlying case decades before anyone built an app for this: American civic life declined not because people stopped meeting new people at any given moment, but because the standing structures that brought the same people back into contact with each other — bowling leagues, union halls, church groups, garden clubs — thinned out. Putnam distinguished bonding social capital, the deep ties within a group you keep returning to, from bridging social capital, the looser ties across different groups. Apps are extremely good at bridging. Almost nothing about their design helps with bonding, because bonding takes repetition, and repetition takes weeks, and weeks don’t show up in a demo.

The neighbour finding, and what it actually shows

AARP’s 2018 survey of adults 45 and older produced a statistic that’s easy to misread. Among people who had spoken to their neighbours, 33% were lonely. Among those who had never spoken to a neighbour, 61% were. It’s tempting to read that as “talking to strangers cures loneliness,” which would support the meet-more-people theory. But a neighbour is not a stranger you meet once. A neighbour is, by definition, someone you are geographically arranged to keep encountering — on the stairs, at the mailbox, walking the same dog past the same window. The finding isn’t about the value of a single conversation. It’s about what happens when your physical life is set up so that a conversation is likely to become five conversations, without you having to plan any of them.

That is the actual mechanism behind most durable friendships: proximity plus repetition, doing the scheduling work for you. Eric Klinenberg’s Palaces for the People makes a related point about physical infrastructure — libraries, parks, senior centres — arguing that these spaces matter less because of what happens on a single visit than because they’re places people return to on a schedule, which is exactly what turns a stranger into someone you recognize, and eventually someone you’d call.

What this means for the apps you actually have

Once you take repetition seriously, the tools split into two rough categories, and it’s worth being honest about which is which.

One-off matching apps connect you with a new person, once, usually around a shared interest or a personality questionnaire. Buffet, which launched in Los Angeles and was covered by TechCrunch in 2024, works this way: you enter interests, answer some personality questions, and an algorithm matches you with one other person and a venue to meet at. Timeleft does something structurally similar for group dining — weekly dinners, small groups matched by personality, no messaging beforehand, so the whole thing happens the night of. Both are honest about what they offer: a lowered barrier to a first meeting. Neither promises, or is designed to produce, a second one. Whether you see that person again is entirely up to what the two of you do afterward, and most matching apps have nothing built in to make that happen. If the meeting goes well and nobody proposes a repeat, the app has done its job and the friendship has not started.

Recurring-structure platforms work the other way around. Meetup, which has been running since 2002 and reports around 60 million members, is organized not around one-time introductions but around groups that meet on a schedule — the same running club, the same board game night, the same book group, week after week. The individual event is less important than the fact that it recurs, and that anyone who shows up twice starts to become a familiar face rather than a stranger. This is a less exciting product to describe, which is probably why it gets talked about less than the flashier matching apps, but it is structurally closer to what actually produces close friends, because it does the one thing matching apps don’t: it puts the repetition into the design instead of leaving it to you.

Neither category is wrong to exist. They solve different problems. The mistake is expecting a one-off match to do the work of a standing group, and then concluding that the whole enterprise of meeting people doesn’t work when a single good conversation doesn’t turn into a friendship on its own.

What to actually do with this

If you’re choosing a tool, ask which category it’s in before you download anything. If it’s a matching app, treat the first meeting as a screening step, not the goal, and decide before you go what you’ll propose if it goes well — a specific thing, on a specific future date, not “we should hang out again.” If it’s a recurring-group platform, the useful move is boring: pick one group, go four times in a row even if the first two are awkward, before deciding whether it’s for you. Four is roughly the point at which people stop being strangers to each other and start being the ones who notice if you don’t show up.

The other option costs nothing and doesn’t require an app at all: find a place you’re already near and build a habit of returning to it at a fixed time. A regular library programme, a run club, a standing seat at the same café on the same morning. Klinenberg’s argument about social infrastructure applies as much to a coffee shop with a regular Tuesday crowd as it does to a library. The venue matters less than the fixed time.

What doesn’t help, on the evidence, is treating each new introduction as a fresh attempt and starting over every time one doesn’t lead anywhere. The problem was never that you hadn’t met enough people. It’s that almost nothing in modern life brings you back to the same ones automatically anymore, so you have to build that back in on purpose — and no algorithm is going to do that part for you.