People Like You More Than You Think

You leave a conversation with someone you just met, and on the walk home you replay it: the pause that went on a beat too long, the joke that did not quite land, the moment you talked too much about yourself. You conclude, quietly, that they were not especially impressed. Meanwhile, in all likelihood, they are walking home thinking almost exactly the same thing about themselves - and rating you considerably more favourably than you rated yourself.

A gap with a name

In 2018, Erica Boothby, Gus Cooney, Gillian Sandstrom and Margaret Clark ran a series of studies putting strangers together for short conversations - in a psychology lab, in workshop settings, sometimes for a few minutes and sometimes over a longer session - and afterward asked each person two things: how much did you like your conversation partner, and how much do you think they liked you? Across every version of the study, a consistent gap appeared. People reliably underestimated how much their partner had liked them and enjoyed the exchange. They called it the liking gap, and it held up regardless of how the conversation had actually gone, how long it lasted, or how anxious the participants said they felt going in.

In 2021, Adam Mastroianni, Cooney, Boothby and Andrew Reece showed the same underestimation extends beyond one-on-one chats: whole groups and teams working together also tend to underrate how much the group as a whole likes each of its members. This is not a quirk of meeting a stranger for five minutes once. It is a durable, general habit of guessing wrong about our own social impact, in directions that always undersell us.

That extension matters practically, because most relationships are not a single conversation - they are dozens of small interactions accumulating over a new job, a shared class, a new circle of friends. If each of those interactions carries its own quiet, private underestimate, the gap does not necessarily fix itself just because you have now spoken to someone five times instead of once. It can just as easily persist, shaping how welcome someone feels in a group that, in reality, already likes them fine.

Why we get our own impact so wrong

Part of the explanation is a related bias documented by Thomas Gilovich, Victoria Medvec and Kenneth Savitsky in 2000: the spotlight effect, our tendency to overestimate how much other people notice and remember about us. In conversation, this compounds neatly. You are intensely aware of every stumble you make, because you are the one making it and it is happening inside your own head in real time. Your conversation partner, meanwhile, is mostly occupied managing their own stumbles, self-conscious about their own pauses and their own perceived awkwardness - and is considerably less focused on cataloguing yours than you assume. The two of you are each, privately, marking yourselves down for mistakes the other barely registered. Each side is, in effect, grading its own performance against a much stricter judge than the one actually in the room.

We get it wrong before the conversation even starts

The miscalculation does not begin only in hindsight. Nicholas Epley and Juliana Schroeder asked commuters to predict how pleasant it would be to strike up a conversation with a stranger on a train or bus, compared with sitting quietly alone. People predicted solitude would be the more pleasant option by a wide margin. When some were then randomly assigned to actually talk to the stranger next to them, they reported enjoying the trip more than the people who had been told to sit in silence - and considerably more than they themselves had predicted beforehand. The underestimate runs in both directions: before a conversation, we lowball how good it will be, and after it, we lowball how well we came across in it.

What this means for you

If you tend to walk away from social situations assuming you were the weaker link, the awkward one, the person who talked too much or too little, the research says your assumption is probably running well below reality, not a fair reading of what actually happened. This is worth keeping in mind if approaching new people, small talk, or speaking up in a group makes you anxious: the outcome you are dreading is, on average, considerably better than the outcome you are predicting, and the harsh replay afterward is a documented bias, not a report from a neutral observer. One practical habit worth borrowing from this research: if you are hesitating to message someone after a conversation that felt fine to you, or holding back from following up because you assume they were not that interested, treat the hesitant assumption itself as the biased one. The base rate the studies describe runs the other way - most people, most of the time, enjoyed the conversation more than you would guess, and are glad rather than put off when you follow up. None of this guarantees any single conversation will go well, and it is not a substitute for support if social anxiety is severe enough to be limiting your life - that is worth raising with a professional rather than working through alone. But for the ordinary, everyday version of this worry, the honest correction runs in a kind direction: you are very likely leaving a better impression than you think, and the people you talked to are probably hoping you will talk to them again.