Dating advice, checked against the research

First message online dating: what gets a reply

Who you write to predicts a reply more than how you word it. What the dating data shows, plus ten first messages built for a short chat.

The Mocha team8 min read

A first message that gets a reply is short, specific and easy to answer in one line, and it gives a little of you alongside the question. That is the part you control. The larger finding, from a study of a month of messaging on one dating site, is that whether you get a reply depends more on who you write to than on how you word it. Most first messages go unanswered, largely for reasons no opener can fix, which is worth knowing before you spend twenty minutes on one.

The reply gradient

Elizabeth Bruch and Mark Newman analyzed a month of messages between users of a popular free American dating site in four cities, New York, Boston, Chicago and Seattle, during January 2014. They built a measure of desirability from who received messages from whom, and then looked at who replied.

Three findings matter for anyone writing a first message.

  • Almost everyone writes up. Men and women contacted people who were, on average, about 25 percent more desirable than themselves by the study's measure. The authors call this the norm, not occasional wishful thinking.
  • Writing up lowers the reply rate. Men were more than twice as likely to hear back from women rated below them as from women rated above them, and for messages sent to more desirable women the reply rate never rose above 21 percent.
  • Most first messages come from men, and women reply selectively. More than 80 percent of first messages in the data were sent by men, and women's average reply rate was below 20 percent. Messages sent by women were more likely to get a reply.

So a man who sends five well written messages to the profiles he most admires, hears nothing, and concludes that his opener is the problem is probably reading the wrong variable. The study's own conclusion is gentler: reaching beyond your league is entirely possible, but it may take messaging two or three times as many people.

Before your message is even read

A second study by Bruch, with Fred Feinberg and Kee Yeun Lee, followed 1,855 people joining a marriage oriented subscription site in the New York area and modeled 1.1 million decisions about whose profile to open and whom to write to. They found that people use deal breakers: cut-offs that remove profiles before any careful comparison happens. Age was the biggest. The most striking one was an omission: both men and women were roughly 20 times less likely to open a profile with no photo, even after the model accounted for everything else.

The practical consequence is that a first message is often judged by someone who has already decided whether you pass a screen. A strong line cannot help a profile that is filtered out, and silence after a good message is not a verdict on the message. If you get no replies at all, fix the photos and the basics before rewriting the opener.

Does a longer first message work better?

People clearly think so. Bruch and Newman found that both men and women wrote substantially longer messages to more desirable partners, up to twice as long in some cases. The payoff was another matter. Among men, only those in Seattle saw better reply rates from longer messages. In all four cities men got slightly fewer replies when their messages were more positively worded, which is a fair warning against openers built from compliments. The authors summarize it plainly: effort put into writing longer or more positive messages may be wasted.

What the conversation studies add

Beyond length and tone, the dating site data says nothing about wording, because nobody tested it. Conversation research fills part of the gap, with the caveat that it studied chats that had already started.

Karen Huang and colleagues at Harvard paired people for 15 minute online chats and told one person to ask either at least nine questions or at most four. Partners of the high question askers liked them more. A later analysis showed the effect ran through follow-up questions, the ones that pick up what the other person has just said, because they signal listening. In a separate set of 110 speed daters on 4 minute dates, people who asked more follow-ups got more second date agreements: the estimate was one extra date across an evening of 20 for an 8 percent difference in follow-up rate. The participants themselves did not predict any of this.

Huang's team add their own caution: asking a string of questions while revealing nothing about yourself could lower liking, and they had not tested where that point lies. A study by Susan Sprecher and colleagues, as summarized in Psychology Today by Susan Krauss Whitbourne, points the same way. Pairs who disclosed back and forth during two 12 minute video calls liked each other more than pairs in which one person talked for 12 minutes and then swapped, and the swapped pairs never caught up.

How to start a conversation on a dating app

Put the findings above in order and they give you a sequence. The order is ours; each step rests on one of the studies.

  1. Pass the screen first. A clear photo and the basics decide whether your profile is opened at all, so fix those before the wording.
  2. Write to more people than feels natural, especially if you tend to write up, because reply rates fall as desirability rises.
  3. Open short and answerable. One line that gives a little of you and asks one thing; in the data, longer and more positive messages did little.
  4. Make your second message a follow-up to their answer rather than a new topic. This is the step with the most direct evidence.
  5. Move it on while it is going well, to a call or a meeting, before the chat runs out of things to say.

If it stalls anyway, our guide to keeping a conversation going has restart messages.

Ten first messages for a short chat

These are our suggestions, written for a conversation with a clock on it. None has been tested against a reply rate, and the notes say where the research stops.

  1. "Hi, I'm Sam. What was the best part of your day so far?" Answerable in a line, with no wrong answer.
  2. "Just got back from pickup basketball and I'm starving. Best thing you've eaten this week?" A small disclosure before the question, the turn taking pattern from the Sprecher study.
  3. "Settle something for me: is a good brunch worth waiting in line for?" An either-or question is easy to join, and the other person can reject the premise.
  4. "What are you looking forward to this week?" Lets them decide how much to say.
  5. "Your photo looks like a market somewhere warm. Where was it?" A profile reference. Reasonable, but no study we read tested it.
  6. "I'll go first: I spent an hour fixing a bike chain tonight. What did your evening look like?" Disclosure, then an open question.
  7. "Planner or see-what-happens person?" Light, and the answer gives you a follow-up.
  8. "What's a place you'd take a friend visiting from out of town?" Gives a thread without asking where they live.
  9. "What are you reading or watching at the moment, if anything?" The "if anything" removes the pressure to impress.
  10. "You said you'd just started climbing. What got you into it?" Not an opener but the message that matters most: the follow-up. This is the one the research supports directly.

Two rest on assumptions the research did not test at all: number 5, because profile references were not studied, and number 3, because humor was not either. What the research gives no support to is the long, effusive opener, which is exactly the one people put the most effort into.

What the evidence cannot tell you

Bruch and Newman studied one free site, one month in 2014, four American cities and heterosexual users only. The screening study used a different site and a single metro area. The conversation studies were laboratory chats and speed dating events, not apps. No study we found compared specific opening lines.

The old advice about depth has limits too. Arthur Aron's well known closeness procedure, which much question based advice descends from, ran for 45 minutes with tasks that escalated gradually, and produced more closeness than comparable small talk. It is not a script for line one. And if a reply is short, remember Erica Boothby's finding that people consistently underestimate how much their conversation partners liked them. A brief answer is not proof that the exchange went badly.

Frequently asked questions

Why do most first messages get no reply?
Mostly for reasons the wording cannot fix. In Bruch and Newman's study of four American cities, women replied to fewer than 20 percent of the messages men sent them, and most messages went to people rated more desirable than the sender, whose reply rates were lowest. A separate study by Bruch, Feinberg and Lee found that many profiles are screened out on age or a missing photo before anyone weighs a message at all.
What is a good opening message for online dating?
A short, specific line that is easy to answer and gives a little of you alongside the question, for example: Long week, finally sitting down with a coffee. What was the best part of yours? No study we found compared opening lines, so that is our suggestion, built on conversation research: back and forth disclosure raised liking in the Sprecher study, and follow-up questions did in Huang's.
Does a longer first message work better?
Not reliably. Bruch and Newman found that people wrote up to twice as long to partners they rated more desirable, yet among men only those in Seattle got a payoff from length, and the authors concluded that effort spent on longer or more positive messages may be wasted. A first line that is easy to answer does more work than a long one.
Should you mention something from their profile?
It is reasonable, but no study we read tested profile references as openers. The nearest evidence is Huang and colleagues' finding that follow-up questions raise liking because they show you listened. A question about something the other person chose to show is the same move made one step earlier, so keep it specific and avoid comments on their body.
Who should send the first message?
Whoever wants the conversation. In Bruch and Newman's data more than 80 percent of first messages came from men, yet messages sent by women were more likely to get a reply. That is one site in 2014 and only heterosexual users, so treat it as a description of that market rather than a rule.

Sources

  1. Bruch and Newman (2018), Aspirational pursuit of mates in online dating markets, Science Advances Full text read October 7, 2026. Supports the four cities (New York, Boston, Chicago, Seattle), a popular free site, the observation month of January 2014, the exclusion of gay and bisexual users, men sending more than 80 percent of first messages, women's average reply rate below 20 percent, people writing to partners about 25 percent more desirable than themselves, a reply rate from more desirable women that never rises above 21 percent, women's messages being more likely to get a reply, longer messages paying off for men only in Seattle, slightly lower reply rates for men who wrote more positively, and the authors' line that extra effort on length or positivity may be wasted. Rechecked October 16, 2026.
  2. Bruch, Feinberg and Lee (2016), Extracting multistage screening rules from online dating activity data, PNAS Full text read October 7, 2026. Supports the 1,855 New York users of a subscription site, the 1.1 million browsing and writing decisions, the two stage model (browse, then write), deal breakers that remove profiles before any detailed comparison, age as the biggest deal breaker, and profiles without a photo being roughly 20 times less likely to be browsed. Rechecked October 16, 2026.
  3. Huang, Yeomans, Brooks, Minson and Gino (2017), It doesn't hurt to ask: question-asking increases liking, Journal of Personality and Social Psychology Full text read October 7, 2026. Supports the 15 minute online chats, the instruction to ask at least nine or at most four questions, higher liking for the high question askers, follow-up questions as the mechanism (responsiveness), the 110 speed daters on 4 minute dates, the estimate that an 8 percent difference in follow-up rate goes with one extra second date over 20, the finding that people do not predict the effect, and the authors' own caution that many questions without self-disclosure might lower liking. Rechecked October 16, 2026.
  4. Whitbourne (2014), The secret to revealing your secrets, Psychology Today, summarizing Sprecher, Treger, Wondra, Hilaire and Wallpe (2013), Taking turns, Journal of Experimental Social Psychology Read October 7, 2026. The primary article (doi 10.1016/j.jesp.2013.03.017) was not read, so the design is cited from this summary by Susan Krauss Whitbourne, PhD: undergraduate pairs, two 12 minute Skype conversations, back and forth disclosure against one person disclosing for 12 minutes and then swapping, higher liking in the back and forth pairs that the swapped pairs never caught up with. Rechecked October 16, 2026.
  5. Boothby, Cooney, Sandstrom and Clark (2018), The liking gap in conversations, Psychological Science (University of Essex repository copy) Abstract and repository full text read October 7, 2026. Supports the finding that after a conversation people systematically underestimate how much their partner liked them and enjoyed their company.
  6. Aron, Melinat, Aron, Vallone and Bator (1997), The experimental generation of interpersonal closeness, Personality and Social Psychology Bulletin Abstract read on October 7, 2026; the full paper was not read. Supports the 45 minute procedure of gradually escalating self-disclosure tasks and greater closeness than comparable small talk in Study 1.