What do you say to start a conversation with a stranger?

Most advice about openers is written as though the opener is the hard part. It isn't. The research points somewhere duller and more useful: what separates people who have these conversations from people who don't is almost never the quality of the first line. It is whether a first line got sent.

The bar is lower than you are setting it

When commuters were asked whether they would strike up a conversation with a stranger, the majority in every setting said no — in one context 93% of them. The reasons people give are consistent: they won't know what to say, or the other person will not want to be spoken to. When the same kind of people were actually assigned to do it, the conversations went better than predicted.

Nobody in those studies was given a clever opener. They were told to start a conversation and find out something about the other person. That was the whole technique, and it was enough.

What we think works in text

Anonymous chat removes most of what makes an opener difficult — there is no room to read, no approach to time, no way to be visibly rejected. What remains is that a message has to give the other person something to answer.

The one that isn't about the opener at all

The strongest evidence here is about what you do after they answer. Across studies of real conversations, people who asked more questions were better liked — and the effect was carried specifically by follow-up questions, the ones that pick up the thing the other person just said. Very few people ask enough of them.

So the useful shape is: open with something answerable, and then, when they answer, ask about what they actually said rather than moving on to your next prepared line. That is most of the skill, and it is one move.

If nothing comes back

Here the research runs out, and it is worth saying so rather than stretching it. Every study above put two people together and started the clock. None of them covers the step that comes first online: being picked.

On a train you talk to whoever is sitting next to you. Here, people work from a roster they can filter — by gender, by age, by country, by language, by what someone says they are up for. A good deal of the deciding is therefore done before anybody has read a word you wrote.

This is the one place where being online is genuinely different rather than merely untested, and it cuts in your favour. The explanation that feels most obvious from the inside — that it was something about you — is the one you have the least evidence for and the least ability to judge. Send another one to somebody else. That is not resilience advice, it is just the only move that produces new information.

What this rests on, and where it doesn't

The commuter studies were spoken and face to face, and the question-asking research covered online chat and face-to-face speed dating — closer to here, but still with a face or a second date attached. Nobody has run these experiments on anonymous text with strangers who will never meet. Stripped of everything we added, what the research supports is narrow: that the reluctance is near-universal, that the reasons people give for it did not survive testing, that no technique was needed to get the result, and that follow-up questions do the work.