Why the second question wins people over

Someone at a party says they’re planning a trip to Canada. You’ve been to Canada. You have a story about a rental car and a snowstorm, and it’s already loading. You tell it, and it goes well. On the way home you notice that you know everything about your trip and nothing about theirs.

Here’s the finding. Strangers who were told to ask more questions in a chat were liked more by their partners, 5.79 against 5.31 on a 7-point scale. The askers had no idea it worked.

A chat with a secret instruction

Karen Huang and four colleagues at Harvard started in a lab (Huang et al. 2017). Strangers sat in separate cubicles and chatted by instant message for 15 minutes. Their only task was to get to know each other. After exclusions, 398 people took part, in 199 pairs.

One person in each pair got a secret instruction. Some were told to ask at least nine questions. Others were told to ask at most four. Their partners didn’t know.

After the chat, the partners rated how much they liked the person they’d talked to. Those who got many questions gave 5.79. Those who got few gave 5.31.

Then the askers guessed how much their partner liked them. The many-question askers guessed 5.27. The few-question askers guessed 5.19. Nearly the same. They had changed how they were seen and didn’t notice.

The team ran the chat again online, then showed those transcripts to 612 new readers. The readers didn’t like the big askers more. They liked the people who got asked more questions. Inside a conversation, you like the person asking you. Watching from outside, you like the person talking.

Which questions counted

Six research assistants sorted every question into types. A follow-up picks up what the partner said a moment ago. A switch question jumps to a new topic. Follow-ups were the most common type, at 40.51% of all questions. They were also the only type that went with liking. People who asked more full-switch questions came across as less responsive.

The paper’s own example of a follow-up comes straight from the data. One person wrote, “I’m planning a trip to Canada.” The other replied, “Oh, cool. Have you ever been there before?”

Four minutes, with microphones

The last study left the lab. The team used recordings from earlier speed-dating research: 110 people in three sessions, each going on 15–19 dates of four minutes. Everyone wore a microphone. After each date, each person marked whether they wanted a second date.

The researchers trained a program to spot follow-up questions in the transcripts. People who asked more follow-ups got more yeses. By the authors’ estimate, an 8% difference in follow-up rate went with about one more second date over an evening of 20 dates. This part is correlational, so it can’t show that the questions caused the yeses.

Why we don’t see it

The authors’ explanation is responsiveness: the sense that someone listens, understands, validates and cares. A follow-up shows it. You can’t ask one without hearing the last sentence.

They also have a guess about why people don’t predict the effect. When we picture a conversation, we picture what we’ll say at one moment, not the back-and-forth. Our own stories are always within reach. A good reply to someone else’s story only exists after they’ve told it.

The question you can’t prepare

The first question can come from anywhere. The second one can only come from listening. There’s a longer guide to follow-up questions on this site.

The rule card in Thesis gives three example questions, and all three are follow-ups: What did that feel like? Who else was there? Has it always meant that to you? See also follow-up questions for managers, and why people like being asked.

This week, when someone mentions a trip, a job or a bad night’s sleep, hold your story for one turn. Ask one thing about theirs. Your story will still be there.