Conversation Starters for a First Date

Conversation starters for dating work when they give the other person something easy to answer while both of you are still deciding whether to stay. That is the constraint on the first twenty minutes of a first date: two adults auditioning each other, both aware of it. An opener that would be neutral at a party reads as screening here.

Key Takeaways:

• Every question on a first date is heard twice, once as a question and once as a criterion, which is why neutral openers feel loaded
• Assigned facts such as work, hometown and siblings produce complete answers, and a complete answer leaves nothing to ask about
• The nearest thing to a date in the question-asking research was a speed-dating round, four minutes a head with a bell on it, where the people asking more follow-ups were picked for a second meeting more often (Huang et al., 2017)
• Dread outruns the event: participants forecast a substantial conversation as worse than it proved, and every one of them had agreed to talk before the forecast was taken (Kardas, Kumar and Epley, 2022)
• Waiting for a date to end itself means waiting for something that rarely happens on cue (Mastroianni, Cooney and colleagues, 2021), so the exit has to be spoken

Why Everything You Ask Is Heard Twice

A question on a first date does two jobs at once. It asks for information, and it announces what you are measuring. “What do you do?” is an ordinary sentence in a hallway and a screening question across a small table, and your date hears both versions before answering either.

That is not paranoia on their part. They are running the same process from the other side, which is why their polite question about your weekend can land like a form to fill in. The same form-filling feeling turns a family table that only meets at holidays into an annual exchange of headlines.

Advice built for parties transfers badly for this reason. In a crowded room a flat opener costs nothing, because either of you can drift off in thirty seconds. Across a table with two hours implied, a flat opener has to be lived with. Neither of you has anywhere to put a bad answer, and that, rather than shyness, produces the stiff first ten minutes people later describe as no chemistry.

There is a way around this that has nothing to do with being charming. Ask about things the person chose instead of things they were assigned. Work and hometown are assignments. What they read twice, what they quit, what they overspend on, the opinion they hold that irritates their friends: those are choices, and a choice arrives with an opinion already attached to it. The answer is retrieval rather than invention, so nobody has to perform. The same choice-over-assignment move is what later revives couples who have run out of things to say.

Before you sit down, pick two things you are honestly curious about that a stranger could answer without preparing. That is the whole preparation, and it beats a memorized list.

The First Twenty Minutes, Roughly in Order

The opening stretch splits into three phases, and each one carries a different amount of weight. Putting the wrong weight on the wrong phase is the most common way a promising evening stalls out.

The first three minutes are logistics and voice. Coats and ordering. Content does not matter yet, and trying to make it matter is the error; the only job is to sound like a person rather than a profile. Say the obvious true thing about where you are: “this place is much louder than its photos,” or “I have never once ordered correctly here.” Either line can be answered in four words by somebody whose hands are full of coat and phone.

Minutes 3 to 10 are where the ladder tempts you, and where one decent question replaces four weak ones:

• “What did you get out of the way today so you could come out tonight?”
• “What is something you used to be sure about and are not any more?”
• “What do you spend money on that other people think is a waste?”
• “Who in your life would be least surprised that you are here tonight?”

By minute 10 the evening can hold something real, and most people wait far too long to try. The 2022 work by Kardas, Kumar and Epley found a forecasting error: people went in braced for more discomfort than the conversation delivered, and guessed low on how much the other side wanted to hear. Paired participants, told a conversation was coming, in most conditions handed the questions. A date sits close to that, since you also agreed in advance to sit down. On a date the error shows up as an hour of commute times, because neither of you wanted to go first.

A practical version of all this: hold your one real question until the drinks have arrived and something small has already gone slightly wrong. Shared minor friction is the cheapest closeness available, and it shows up by minute ten almost every time.

The Interview Ladder, and What Happens on Rung Four

The ladder is job, hometown, siblings, how long you have lived here, what you do for fun. It fails for one structural reason: every rung has a complete answer. Ask where someone grew up and they tell you. Done. The escape is the same one behind any interesting conversation starter: ask about something with no complete answer.

Climb four rungs and you have a file rather than a conversation: employer, home state, sibling count, gym schedule. Both people feel it happening and neither can name it, so it gets filed under bad chemistry and the evening ends at ninety minutes with a handshake.

The opposite failure is rarer and worse. Someone reads about deep questions, arrives armed, and at minute two asks about your greatest regret, leaving a stranger to either open a vein or deflect before the food arrives.

Those 36 questions get imported into dates constantly, and the import throws away the part that made them work. Aron and colleagues built that procedure in 1997 as a structured, mutual, escalating exchange where both people knew what was happening and both were held to it. A date carries no such agreement. Asking a stranger about their mother’s death over the first drink is not the experiment. It is just a stranger asking about a dead mother.

If you catch yourself on rung three, stop asking and say something instead. A statement with an opinion inside it gives your date something to push against, and pushing back is far easier than answering.

Conversation Starters for Dating That Survive a Nervous Answer

A good question will often come back compressed, because your date is nervous, and what you get is seven words. The opener did not fail. What decides the next five minutes is what you do with those seven words, and almost nobody does the useful thing.

Huang and colleagues put numbers on this in 2017: in a speed-dating round of four minutes a head, the people asking more follow-ups got picked for a second meeting more often. Speed dating is not a first date. Everyone there had signed up to be assessed on a clock, a blunter agreement than yours over text. The back-and-forth that got you to the table is its own skill, starting a conversation by text with someone new, and it does not transfer cleanly to sitting across from them. What crosses over is the ranking: the second question outperformed the first.

Their compressed answer will contain at least one noun you did not put there. Ask about that noun. “I moved here for work, it was fine” gives you two: the move, and the word fine. Either of these gets you a real answer: “Fine is doing some heavy lifting in that sentence.” Or: “What did you leave to come here?”

Openers that hold up under nerves share one property: they can be answered badly and still leave material behind.

• “What is your read on this place? I picked it off two photos.”
• “What did your week look like, honestly, not the polite version?”
• “What is something you are stubborn about?”
• “What do people get wrong about your job?”

When You Already Know It Is a No

Decide slowly and leave clearly. Those are different skills, and most people invert them: they decide at minute six and then sit there for another hour hoping the evening ends itself.

The deciding part deserves patience, because early reads are unreliable in a specific direction. Boothby and colleagues reported in 2018 that people finish a conversation having marked down what their partner made of them. Read the setup, though: strangers paired off by an experimenter, nothing riding on it, nobody deciding whether to see anyone again. A date inverts all three. The direction of the error survives, and it beats your own verdict at minute six: nerves compress a person the way boredom does. A quiet forty minutes and a bad forty minutes look identical from outside.

The leaving part deserves none of that patience. In 2021 Mastroianni, Cooney and colleagues checked when conversations actually stopped against when the people in them had wanted to, and the two almost never lined up. Some came from lab pairs who had agreed to talk, some from people recalling conversations with someone they knew. None from a date. But if the stopping point is that unreliable between volunteers and old friends, waiting for one across a table from a stranger is waiting for something that mostly does not arrive. Finish the drink, be decent, and say the exit out loud: “I am going to head off, but thank you for this.” Complete sentence, no apology, nobody has ever been wounded by it. Saying it cleanly is a small version of preparing for a hard conversation before you have it.

One limit worth naming plainly: none of this applies if you feel unsafe. Leaving abruptly is always available and never requires a script.

FAQ

What do I talk about on a first date after the small talk runs out? Ask about something the other person chose rather than something they were assigned. Their job, hometown and family are assignments and have complete answers; what they quit, what they overspend on, and what they have changed their mind about are choices, and each one comes with an opinion already formed. One choice question typically buys ten minutes.

How do I start a conversation on a first date without sounding like an interview? Alternate a question with a statement that contains an opinion. An interview is a string of questions with no disclosure between them, so the fix is not softer questions, it is putting something of your own on the table every second turn. Answering your own question first works well when the table has gone quiet.

Is it too soon to ask a deep question on a first date? Around minute 10 is not too soon, and most people wait much longer than they need to. The 2022 experiments point one way: dread ran ahead of the actual conversation, among pairs who had already agreed to have one (Kardas, Kumar and Epley, 2022), which describes a date fairly well. Ask about a decision rather than a wound.

What do I do if they barely talk? Treat short answers as compression rather than disinterest for the first twenty minutes, and mine each answer for one noun to ask about. If three follow-ups in a row still come back in five words, the read is fair and you can end early without guilt. Nerves and boredom look the same from the outside, which is the entire problem.

A first date is not a test you can study for. It is a conversation with an unusually short warm-up and two people watching themselves have it, and nothing you memorize survives contact with a nervous stranger. What survives is the habit of asking about the last thing they said, offering something of your own before asking for more, and going slightly deeper twenty minutes sooner than feels comfortable. Get those right and the specific opener stops mattering, which it never did much. If you want to wear in the reflex of asking about the thing someone chose before you are sitting across from anyone, Lona makes a decent private place to rehearse it: an AI you talk to on your own schedule that keeps the through-line across days. It is practice, not a date. And no substitute for actually going on one.

Sources

• Huang, K. et al., “It Doesn’t Hurt to Ask: Question-Asking Increases Liking,” Journal of Personality and Social Psychology, 2017
• Boothby, E. J. et al., “The Liking Gap in Conversations,” Psychological Science, 2018
• Kardas, M., Kumar, A. & Epley, N., “Overly Shallow?: Miscalibrated Expectations Create a Barrier to Deeper Conversation,” JPSP, 2022
• Mastroianni, A. M., Cooney, G. et al., “Do Conversations End When People Want Them To?”, PNAS, 2021
• Aron, A. et al., “The Experimental Generation of Interpersonal Closeness,” 1997