Reporting on loneliness and how people find their way out of itTools & apps
Be Social.

The Science

What Texting Someone Isn't Doing

Online contact can maintain a friendship, but it rarely builds one, and it does not replace the physical, unplanned run-ins that used to make up most of social life.


Photograph · Pexels

You texted three people back today. You liked a friend’s post, replied to a group chat, maybe had a video call with someone who moved away. By most measures, that’s a full social day. And yet the evening still feels thin.

There’s a reason for that, and it isn’t that you did contact wrong. It’s that contact and connection are two different things, and online tools are much better at the first one than the second.

Closeness matters more than volume

A nationally representative study published this month looked at something most surveys skip: not how much time people spend on social media, but how close they actually are to the people they’re in touch with there. The finding was not “more online contact equals less loneliness” or the reverse. It was that closeness moderates the relationship. Scrolling past acquaintances and keeping up with close friends are not the same activity wearing the same interface.

This matches something people already sense but rarely say out loud: a like from someone you barely know does nothing, and a proper conversation with someone who knows your history does something real, and the fact that both happen inside the same app is a coincidence of design, not a sign they’re equivalent.

What online contact is actually good for

It is a genuinely good maintenance tool. If you have an existing close friendship and you move cities, or one of you has a baby, or work swallows your evenings, a text thread keeps that friendship alive between meetings in a way that would have been much harder before phones existed. Robert Putnam’s account of declining civic life in Bowling Alone was written before smartphones, and one of the few things that has plausibly improved since is exactly this: distance no longer kills a friendship the way it once did.

But maintenance is not the same job as formation. The Harvard Making Caring Common report on loneliness in America found that about half of lonely young adults said no one had taken more than a few minutes recently to ask how they were doing in a way that felt genuine. That’s not a supply problem in the sense of not enough messages arriving. People get plenty of messages. It’s a depth problem. A check-in text can be sent in four seconds and read the same way. The kind of exchange that actually reduces loneliness tends to take longer than that, and tends to require being in the same room.

The thing online contact cannot do

Here is the part that’s easy to miss: a huge amount of the connection that used to make up ordinary social life didn’t come from conversations at all. It came from being physically present around other people, repeatedly, without a specific agenda. Eric Klinenberg’s research on social infrastructure describes this as the work done by libraries, parks, diners, and barbershops — places where you see the same faces often enough that they become familiar, and familiarity is most of what makes a place feel like it’s yours.

Research published in 2025 tracked twelve categories of these places — coffee shops, libraries, museums, recreation centres, restaurants — across the United States between 2019 and 2021, and found closures in every category, concentrated most heavily in poorer and rural areas. A commentary on that research from the University of Colorado Boulder framed the loss plainly: fewer senior centres and community spaces is a cause of loneliness, not just a symptom of it. No app replaces a place. You cannot have an unplanned run-in with a text thread. You cannot be recognised by a group chat when you walk in the door.

This is the substitution that doesn’t work, no matter how good the messaging is. Online contact can carry a relationship that already has a physical foundation. It struggles to build that foundation from nothing.

Where it gets more complicated

The newest, and more unsettling, version of this question is what happens when the “contact” isn’t with a person at all. A twelve-month study following more than two thousand adults across four countries found that increased use of AI companionship predicted increased loneliness over time, not the other way around. It isn’t simply that lonely people reach for chatbots, which they do — it’s that reaching for them seems to deepen the loneliness rather than relieve it. Reporting on the same body of research describes short-term comfort paired with longer-term distress, with heavy daily use correlated with reduced human social engagement. The American Psychological Association’s review of this area is more cautious, noting the effects depend on the person, the design, and how much they use it — but even that caution is itself informative. Nobody in this literature is claiming an AI companion substitutes for a friend. The most optimistic finding available is “it depends,” which is a much weaker claim than the marketing around these tools tends to make.

What to actually do with this

None of this is an argument to quit texting people. It’s an argument to notice what texting is for. Use it to keep a real friendship warm across a week you can’t see them — that’s its actual job, and it does that job well. Don’t expect it to build a new friendship from a cold start, and don’t let a satisfying group chat replace the effort of getting a body into a room with other bodies on a regular basis.

A study of college students published last year, looking at what actually predicted wellbeing as students came out of the pandemic, found the connections that mattered were the ones with some regularity and depth to them — not sheer number of contacts. If you want the evening to stop feeling thin, the fix probably isn’t a longer message. It’s a standing Tuesday something, at a place, with the same few people, often enough that you stop having to plan it.

Where this comes from