Think back to the last real conversation you had — not logistics, not the weather, an actual one. Now try to recall a single question the other person asked you. Not a courtesy how are you lobbed at the start like a coin into a fountain. A real question, one that required them to have listened to your previous sentence.
Most people can't. And here's the part that stings: the other person almost certainly can't recall one from you either.
We spend our conversations waiting. Assembling. Watching for the gap where our anecdote fits. We are so busy managing the impression we're making that we forget the single cheapest, best-documented way to make a good one — and it's not the anecdote. It's the question.
The finding nobody acts on
In a 2017 paper in the Journal of Personality and Social Psychology, researchers led by Karen Huang and Alison Wood Brooks at Harvard did something almost embarrassingly simple. They took hundreds of conversations — some in a lab, over text chat; some at a real speed-dating event — and counted the questions.
Then they looked at who got liked.
The people who asked more questions were liked more. Not marginally, and not because they were charming or clever — the effect held when the researchers instructed people to ask more questions, meaning it wasn't just that likable people happen to be curious. Causation ran the right direction. And the effect concentrated in one specific type: the follow-up question. The one that proves you were listening to the last thing they said rather than rummaging in your own head for what to say next.
At the speed-dating event, people who asked more follow-up questions were more likely to be asked on a second date.
Here is the strange part. In another arm of the research, when people were asked to predict whether question-asking would make them better liked, they didn't think it would. The behavior that reliably works is the behavior we don't believe in.
Why we don't ask
So why does something this effective, this free, and this available in every conversation you will have this week feel so unnatural to do?
The honest answer is that we are not primarily trying to connect. We are trying not to look bad.
The conversational self is an anxious creature. It believes the conversation is an audition, and that the way to pass an audition is to perform. Asking a question, in this framing, is a forfeited turn — you had the floor, you had a chance to be interesting, and you spent it pointing the light at someone else. Worse, a question can expose you. You didn't know that? Every question carries a small confession of ignorance, and the anxious self will pay a lot to avoid confessing anything.
This is compounded by a second, well-documented bias. Erica Boothby, Gus Cooney, and colleagues have described what they call the liking gap: after a conversation with someone new, people systematically underestimate how much the other person liked them. We walk away replaying our own awkward pauses, unaware that the other person is doing exactly the same thing and remembers none of ours. Their research suggests this gap can persist — not just for a few minutes, but across acquaintanceships that last months.
So picture what this does inside a conversation. You are convinced you're being judged more harshly than you are. You respond by performing harder — talking more, self-presenting, managing. Which means asking less. Which means the one behavior that would actually raise how much they like you is the one your anxiety systematically suppresses.
The fear of seeming uninteresting makes us uninteresting.
What a question actually does
When someone asks you a good question, three things happen at once, and none of them are about information.
First, it is proof of attention. A follow-up question cannot be faked in advance. It's a receipt showing you processed what was said. Nicholas Epley's research on conversation suggests much of the warmth we feel from being talked to comes not from the content but from the evidence of a mind engaging with ours.
Second, it hands the other person the pleasure of self-disclosure. Diana Tamir and Jason Mitchell at Harvard put people in an fMRI scanner and found that disclosing information about oneself engaged brain regions associated with reward — the same dopaminergic circuitry that responds to food and money. People will even forgo small amounts of money for the chance to talk about themselves. When you ask a question, you are not being self-effacing. You are handing over something people genuinely want.
Third — and this is the one that reorganizes how you think about conversation — asking a question makes the asker more interested. Curiosity isn't a mood that descends on you and then licenses questions. It runs the other way just as often. Ask about someone's job and you will find, three sentences in, that you have located the one weird thing about it. The question generates the interest, not the reverse. Which means the person who complains that other people are boring has usually just never asked them anything.
The specific skill
There's a hierarchy here, and most conversational advice ignores it. Huang's team distinguished between question types, and they are not equal.
Mirror questions — how was your weekend? / good, how was yours? — are politeness. They cost nothing and signal nothing.
Switch questions change the topic. They're often a graceful way of saying I would like to stop talking about this.
Full-switch questions introduce something entirely new — usually because the asker was never listening at all.
Follow-up questions are the ones that carry the weight. They take the last thing said and go one level down. They're the only kind that require you to have actually been present, and they're the kind that predicted second dates.
The skill, then, is narrower than "be curious." It's this: notice the door. In almost every sentence a person says, there is a door — a clause they said quickly, an adjective they chose oddly, a thing they mentioned and moved past. "I moved back here after a rough couple of years." "We ended up not doing it." "It's fine, it's a job." Those are doors. Most of us walk past them out of a politeness that is really just self-absorption.
Your next moves
- In your next conversation, set a floor of three follow-up questions. Not questions — follow-up questions, each one built from a word the other person just used. Count them on your fingers under the table if you have to. You will be shocked how hard three is.
- Practice the phrase "say more about that." It is the lowest-effort door-opener in the language, it costs you nothing, and it works on almost any sentence. Use it once today.
- Kill one anecdote. The next time you notice yourself waiting for a gap to insert a story you've told before, ask something instead. Notice what the conversation does. Notice, afterward, whether you feel worse.
- At dinner tonight, ask one person a question you've never asked them. The people we know best get the fewest questions, because we assume the file is complete. It isn't. Ask a partner or parent what they were most afraid of at your age.
- Write down the questions you didn't ask. After a conversation, jot the doors you walked past — the thing they said about their brother, the job they left. These are the openings for next time, and they evaporate within the hour.
The larger reframe
The reason all of this matters isn't tactical. It's that we have quietly accepted a model of ourselves where being interesting is a property we possess or lack, a fixed quantity we bring to the room and hope is enough. It isn't. It's a thing that happens between two people when one of them decides to be genuinely, unstrategically interested in the other. The likability wasn't in the answer you gave. It was in the question you asked.
Curiosity, it turns out, is not a personality trait. It's a practice — and like any practice, it survives only if you capture it. That's why we built CurioKit: a place to hold the questions you'd otherwise lose — the ones you meant to ask someone, the ones that occurred to you at 11pm, the doors you walked past. It turns fleeting curiosity into something you can return to. If you'd like to stop letting your best questions evaporate, you can find it at curiokit.lumenlabs.works.