There is a move available in almost every conversation that costs nothing, requires no wit, and reliably makes people like you more. It is not a clever opener or a good story. It is the follow-up question — the second question, the one that shows you actually listened to the answer to the first.
In a speed-dating study, participants in the top third for question-asking secured second dates 39% of the time. Those in the bottom third managed 22%. Follow-up questions specifically were the strongest predictor of being liked.
Why follow-up questions work when other questions do not
Not all questions are equal. “What do you do?” is a question. So is “Where are you from?” Neither does much, because either could have been asked of anyone. They gather information without demonstrating attention.
A follow-up question cannot be pre-planned. It can only be constructed from what the other person just said, which makes it evidence — not a claim, but proof — that you were listening. That is the mechanism researchers point to: it is not politeness, it is verification.
Research on high-quality listening supports this. Across two pre-registered studies of conversations between strangers, observed high-quality listening behaviours were linked to both behavioural and subjective markers of social connection. Participants asked to display those behaviours — asking more follow-up questions among them — produced measurably more connection with their partners.
The finding is unglamorous. Being interesting is less effective than being interested, and the follow-up question is the cheapest available demonstration of interest.
The awkwardness gap
A separate line of research explains why people avoid the conversations where this matters most.
Across 12 experiments involving more than 1,800 participants, people consistently predicted that deeper conversations with strangers would feel awkward. They then rated those conversations as significantly more enjoyable and more connecting than the small-talk comparison. The gap between predicted and actual awkwardness produced an effect size of Cohen’s d = 2.23 — enormous by the standards of psychological research, where 0.8 is conventionally considered large.
People are not slightly miscalibrated about this. They are dramatically wrong, in the same direction, over and over.
The practical consequence is a systematic bias toward shallow conversation, driven by a fear that the research says does not materialise. Our piece on the underrated benefits of small talk covers the other side of this: small talk is more valuable than its reputation suggests, even as people overrate how necessary it is to stay there.
The backdrop: we may simply be talking less
These findings land at a moment when researchers are asking a broader question — whether the total amount of spoken conversation in daily life is shrinking as more communication shifts to text.
The concern is not that written communication is inferior for every purpose; it plainly is not. It is that speech carries a bundle of things text strips out: tone of voice, timing, the immediate emotional feedback of hearing how something landed, and the unplanned turns that produce follow-up questions in the first place. A typed exchange can be edited into something smoother than speech and still transmit less.
Talk-time balance is part of this too. Analyses of video-chat conversations have examined how speaking time gets shared between participants and how that distribution shifts as a conversation develops — a dimension that is invisible in text, where there is no real-time competition for the floor.
If speech is genuinely declining as a share of how people communicate, then the skills in this article get less practice than they used to. That makes the follow-up question both more valuable and rarer — a reasonable inference, though not one these studies set out to test.
What these studies do not establish
Several caveats are worth keeping.
The speed-dating figures are correlational. People who ask more follow-up questions get more second dates, but question-asking may be a marker of underlying social skill rather than its active ingredient. Someone attentive enough to ask good follow-ups is likely doing several other things well simultaneously.
Much of this work involves conversations between strangers in controlled settings — laboratory studies and speed-dating events. Whether the same effects hold with a colleague you have known for a decade, or a partner, is a reasonable question the research does not answer.
There is also a dosage problem nobody has mapped. The studies establish that more follow-up questions are generally better; they do not identify where that curve turns. Sustained questioning without reciprocal disclosure becomes an interrogation, and most people can recall a conversation that felt like one.
Finally, laboratory participants instructed to listen well know they are being observed. That is not quite the same as spontaneous behaviour.
How to actually do this
The research points to a few concrete behaviours rather than a general instruction to be a better listener.
- Ask a second question about the same thing. The default move is to answer with your own equivalent story. Resist once and ask instead.
- Follow the specific, not the general. If someone mentions they have just moved back from three years abroad, ask about the moving back, not about the country. The specific detail is what they are actually thinking about.
- Use their words. Echoing a phrase they chose signals you registered the phrasing, not just the gist.
- Let the pause happen. Filling every gap prevents the elaboration that would have given you something to ask about.
- Go deeper sooner than feels comfortable. Given the size of the awkwardness gap, your instinct about what is too personal is probably calibrated too conservatively.
- Reciprocate. Questions without any disclosure of your own tip into interview territory.
If you want structured practice, our conversation starters for couples work as raw material — though on this evidence, what you ask second matters more than what you ask first.
Frequently asked questions
What exactly counts as a follow-up question?
A question that could only have been asked because of what the other person just said. “How long were you there?” after someone mentions living abroad is a follow-up; “what do you do?” is not, because it could have been asked of anyone.
Can you ask too many questions?
Almost certainly, though the research does not identify the threshold. The studied risk is under-asking, not over-asking — but questions without reciprocal disclosure tend to feel like an interview.
Does this work in established relationships?
Most of the research examined strangers, so extending it to long-term relationships is an inference rather than a finding. The underlying mechanism — demonstrating attention — is not obviously limited to new acquaintances.
Why do people avoid deeper conversations if they enjoy them more?
They expect awkwardness that does not arrive. Across 12 experiments people substantially overestimated how uncomfortable deeper conversation would be, and the prediction rather than the experience governs the choice.
Is asking questions better than telling good stories?
For being liked, the evidence favours question-asking. The speed-dating data links question frequency to second dates; there is no comparable finding for storytelling ability.
The short version
The hardest thing about conversation is usually imagined to be knowing what to say. The research suggests the constraint lies elsewhere: people fail to ask a second question, and they avoid depth because they expect an awkwardness that does not show up.
Both are cheap to fix, and neither requires being more interesting than you already are. For a related finding on choosing the right medium, see why you should call instead of texting during a disagreement.
Sources: Communications Psychology on high-quality listening and the American Psychological Association on expected versus actual conversation awkwardness.