How to Talk to Strangers—and Why It Usually Goes Better Than We Expect

The awkward part of talking to a stranger is often the moment before anyone speaks. Research suggests we overestimate rejection, underestimate how much we may enjoy the exchange, and routinely leave small moments of connection unused.

There is a familiar choreography to waiting in public now. Someone takes out a phone. Someone else puts in earbuds. A third person studies a menu they have already read twice. I have done all three, sometimes with impressive efficiency.

The interesting part is that none of this necessarily means we dislike other people. Sometimes we are tired, busy or perfectly happy to be left alone. But sometimes the phone is doing another job: giving us somewhere safe to put our attention while we avoid deciding whether the person beside us might be worth talking to.

That decision is surprisingly difficult because we have very little information. Will they think I am bothering them? Will I have anything to say after the first sentence? What if I make an observation about the weather and discover, twelve seconds later, that I have exhausted my entire conversational repertoire?

Research suggests our predictions are frequently more pessimistic than the conversations themselves.

The Hardest Part May Be the Moment Before Hello

In a well-known series of experiments, researchers Nicholas Epley and Juliana Schroeder asked commuters on trains and buses either to talk with a stranger, remain disconnected or commute as they normally would. People who connected with someone reported a more positive experience than those who remained alone—and separate participants had predicted the opposite. Part of the problem was that people underestimated how interested other passengers would be in talking with them.

Later research on London-area train commuters sharpened the point. Participants understood that a conversation itself might be enjoyable; what they were particularly wary of was trying to start one. They expected initiating contact to be less pleasant than the conversation ultimately proved to be, in part because they anticipated that the stranger would not be interested.

That feels much closer to real life.

Most of us are not sitting on trains thinking, Human conversation: absolutely intolerable. We are thinking, He has headphones. She looks busy. I don’t want to be that person. Maybe I’ll look at my email again.

The barrier is often not conversation.

It is permission.

We Are Poor Judges of How Badly This Is Going to Go

Gillian Sandstrom and Erica Boothby pooled findings from seven studies involving 2,304 people and compared what participants feared before conversations with strangers with what they reported afterward. People worried about not enjoying themselves, not knowing what to say, being disliked and discovering that the other person was not enjoying the conversation either. Across the studies, those fears tended to be exaggerated: the conversations generally went better than expected.

Then there is what psychologists call the liking gap. In a separate series of studies, people who had just met systematically underestimated how much their conversation partners liked them and enjoyed their company. The effect appeared in laboratory conversations between strangers and in people forming new relationships outside the laboratory.

This may explain the peculiar post-conversation autopsy many of us conduct.

Why did I say that?

I talked too much.

That joke wasn’t funny.

Meanwhile, the other person may be walking away thinking you were perfectly pleasant and wondering whether they talked too much.

Apparently, everyone leaves the scene and prosecutes themselves separately.

We May Also Underestimate What a Stranger Has to Offer

There is another mistake hidden in our reluctance to talk: we may assume the conversation will not be particularly interesting.

Across seven experiments published in Proceedings of the National Academy of Sciences, people systematically underestimated how much they would learn from conversations with others. The effect persisted whether participants were given conversation prompts and whether they had been specifically instructed to try to learn something. The uncertainty of not knowing where a conversation might go appears to make us undervalue it before it starts.

That is worth remembering because small talk has terrible public relations.

“What do you think of this rain?” is unlikely to become the sentence your biographer leads with. But small talk is often not the destination; it is how two people establish, very quickly, whether there is somewhere else to go.

The weather may simply be the door handle.

Look for Permission Before You Look for an Opening Line

A particularly useful 2025 field experiment suggests that one obstacle to stranger interaction is wonderfully basic: we do not know whether the other person wants to talk either.

Researchers placed adults in a university-campus setting where a stranger could explicitly signal either openness to conversation or a preference for solitude. Among 563 participants, 56 percent initiated a new connection when the stranger signaled willingness to socialize, compared with 41 percent when no explicit signal was given and 20 percent when the stranger signaled a preference to be left alone. Among participants who themselves had said they were open to meeting someone new, the corresponding figures were 68, 50 and 25 percent.

It was one experiment, in a carefully constructed university setting, so we should not turn those percentages into laws governing every café and bus stop. But the broader idea is useful: people respond to signs of receptiveness.

That means the first skill in talking to strangers may not be speaking at all.

Look up.

Is there eye contact? A smile? An unhurried moment? Has the person already made some little acknowledgment of you? Or are they wearing headphones, buried in a book, typing furiously, avoiding eye contact or otherwise communicating a perfectly reasonable desire to be left alone?

The goal is not to become the person who interprets every human being within six metres as an invitation.

Social confidence without social judgment is just another problem.

You Do Not Need a Clever Opening Line

Once someone seems receptive, I would resist the temptation to become interesting on command.

The easiest conversations usually have something already available to talk about: the place, the delay, the dog, the view, the event you are both attending, the inexplicably complicated ticket machine you are both losing an argument with.

A comment can be enough:

“I wasn’t expecting it to be this busy.”

Or a simple question:

“Have you been here before?”

The usefulness of shared context is that nobody has to wonder why you are speaking. You are both standing inside the answer.

From there, let the other person’s response determine whether there is a conversation. A brief answer accompanied by a return to the phone is probably the end. A longer answer, a smile or a question back is an invitation to continue.

You do not have to rescue every exchange from dying naturally.

Sometimes “Have a good day” is the successful conclusion.

Being Interested Beats Trying to Be Interesting

Once a conversation has started, the skill we tend to worry about most—having impressive things to say—may be less important than what happens while the other person is speaking.

Research published in 2025 examined 646 people in conversations between strangers, including both longer semi-structured conversations and brief small-talk encounters. Behaviours associated with higher-quality listening—such as follow-up questions and verbal acknowledgment—were linked with several measures of social connection. In one of the studies, an intervention that increased follow-up questions was also associated with greater connection between conversation partners.

That gives us wonderfully ordinary advice.

If someone says they are visiting from Porto, you don’t immediately need an amusing Porto story of your own.

Ask what brought them here.

If they mention changing jobs, ask how the new one is going.

If they say their dog is 14, perhaps ask his name before explaining the dietary history of every dog you have owned since 1987.

A good follow-up question communicates something remarkably powerful for a sentence that usually contains fewer than ten words:

I heard what you just said.

That is conversation.

A Conversation Does Not Need to Become a Friendship to Count

One problem with our language around connection is that we tend to reserve its importance for close relationships. Family matters. Friends matter. Partners matter. Of course they do.

But the people at the edges of our lives may matter too.

Sandstrom and Elizabeth Dunn studied what psychologists sometimes call weak ties—classmates, acquaintances and other peripheral relationships. People reported greater happiness and belonging on days when they interacted with more classmates than usual, and across additional studies, interactions with weak ties were associated with social and emotional well-being.

A stranger is not automatically a weak tie; frequently they remain a stranger forever. But today’s person at the coffee counter can become next month’s familiar face, and some of the texture of ordinary community life comes from people whose surnames we may never learn.

The woman at the bakery who knows what you order.

The man who walks his dog when you walk yours.

The receptionist you always joke with.

The neighbour whose conversations rarely travel beyond the weather but somehow make the street feel less anonymous.

Not every meaningful social interaction needs to qualify for the wedding guest list.

If You Feel Rusty, Practice May Matter More Than Advice

Sandstrom, Boothby and Gus Cooney tested this quite directly with a week-long intervention that repeatedly asked participants to find, approach and talk with strangers. As the week progressed, participants became less pessimistic about rejection, felt more confident in their conversational ability and reported less awkwardness. Some of those improvements were still present at a follow-up one week after the intervention ended.

The interesting part is that the expectations changed gradually through experience.

You could read fifty articles telling you that strangers probably won’t reject you. Your nervous system may remain unconvinced.

Then you say hello to the person beside you, nothing terrible happens, and the prediction has acquired some inconvenient new evidence.

Do that repeatedly and the story you tell yourself may begin to change.

What I Wish I’d Known Earlier: confidence does not necessarily arrive before the conversation. Sometimes the conversation is how you acquire it.

Try Contact Before You Try Conversation

If talking to strangers feels unnatural, I wouldn’t begin by setting yourself the task of having fascinating conversations with seven new people before Friday.

Start one level lower.

Make eye contact and say good morning to someone you regularly pass. Thank the person behind the counter without already looking at your phone. Make one observation when you are both waiting for the same thing. Ask one question when the situation offers an obvious one.

Then notice what comes back.

Some people will give you a smile. Some will talk. Some will offer a polite answer and return to whatever they were doing. A few may not respond at all.

None of those outcomes requires a committee meeting afterward.

The skill is partly learning that an interaction can be tiny, mildly awkward and still perfectly successful.

Not Every Silence Needs Fixing

There is an important counterweight to all of this.

Talking to strangers is not universally desirable, and the research itself suggests that people’s own desire for connection matters. In the 2025 signaling experiment, the benefits of explicit openness were driven largely by people who already wanted to make a connection; those who preferred solitude did not suddenly become social butterflies because somebody gave them permission to talk.

Context matters too. Personal safety matters. Cultural norms matter. A person working a customer-facing job is not necessarily inviting personal conversation because they are being friendly. Someone trapped beside you on an airplane cannot conveniently leave. Headphones are often headphones, not a puzzle awaiting your solution.

The aim is not talk to more people at all costs.

It is to become a little less likely to assume that every stranger wants distance—and a little better at recognizing when two people might both be open to something else.