How to flirt over text: examples, emojis and timing
Match their tone, be specific, use one emoji. Before and after text examples, what emoji research says, and what the timing evidence supports.
Flirting over text works when you match the other person's tone, make one playful or specific remark that is easy to answer, and use at most one emoji. The strongest evidence is about matching. In a study of 40 speed dates, 33.3% of pairs whose use of function words (the little words such as "I", "the" and "but") was more alike than the median both wanted future contact, against 9.1% of the rest, and in 86 couples' instant messages the same matching predicted who was still together three months later. You cannot fake that with a template, but it is a good reason to read how someone writes before you reply.
Match their length and mood
Language style matching is measured, not performed: Ireland and colleagues counted function words in natural speed dating transcripts and instant messages, not in an exercise where people copied each other. The useful reading for texting is simple. If they write three words and no punctuation, a paragraph with four exclamation marks will feel like a different conversation. If they write a long, funny message, "haha nice" is a step back.
Matching does not mean mirroring everything. It means staying near their length, their punctuation and their level of silliness, and adding one thing of your own.
Before and after
The usual advice is a list of lines. A rewrite shows what changes.
| They send | Flat reply | Better reply | What changed |
|---|---|---|---|
| "Just got back from the gym, wiped out" | "Nice, what are you up to later?" | "Wiped out, or the smug kind of wiped out?" | Playful, and it asks for a small answer |
| A photo of their dog on the couch | "Cute 😍" | "He looks like he's already won the argument about who gets the couch" | A specific detail, not a compliment |
| "Long day, I'll text you tomorrow" | "Ok" | "Go put your feet up. I'll be here, probably losing at the crossword." | Warm, and it does not demand anything |
| "Good to meet you last night!" | "You too!" | "I enjoyed it more than I expected to, and I had high hopes." | More personal, one notch warmer |
| You wrote a long message, they replied with "haha" | Three more messages | Wait, then a single new topic | No pressure; their reply was an answer |
The last row matters. If someone is replying in fewer words each time, the signal is to match that, not to compensate with more of yours.
Timing
We found no study of reply speed in flirty texting as such. The nearest is an experiment by Lars Teichmann and colleagues at Leuphana University, published in 2026, with 543 participants, on the first text after a first date. Texting the next morning produced the highest relationship intentions, ahead of texting straight after the date or two days later, and the analysis pointed to perceived reciprocity and reliability. It was an experimental scenario about one particular message, not a record of real dates, so it says nothing about how fast to answer in a lively back and forth. What it does suggest is that neither instant nor strategic delay is the safe default for that first message after meeting. Our page on double texting covers sending a second message before you get a reply.
Emojis: how to use them
We found no survey of how American or British daters read particular emoji, so the readings in the table below are our judgment.
Two facts are sound. First, Gesselman, Ta and Garcia found, across two studies, that emoji use with potential partners was associated with keeping a connection going beyond the first date and with more romantic and sexual interactions over the previous year. That is an association and the authors do not claim cause. Second, emoji are ambiguous. Miller and colleagues found that people disagreed about whether the same emoji rendering was positive, neutral or negative 25% of the time, and that platform differences increased the disagreement. In a later study of 2,482 participants, seeing the emoji inside a sentence did not reduce the potential for misreading either.
So the advice is one emoji, after words, chosen from the ones whose standard name tells you what it does. The names below are from the Unicode Consortium's emoji list, version 18.0.
| Emoji | Unicode name | How it tends to read |
|---|---|---|
| 😉 | winking face | Teasing; "I know what I said" |
| 😏 | smirking face | Cheekier than the wink, and easy to overdo |
| 😊 | smiling face with smiling eyes | Warm and safe, not flirty by itself |
| 🥰 | smiling face with hearts | Affectionate; usually reads as more than a first week |
| 😘 | face blowing a kiss | A kiss; fine after a good date, forward before one |
| 🙈 | see-no-evil monkey | Embarrassed or self-aware about what you just said |
Leave 🍆 (Unicode name: eggplant, which British people call an aubergine) and 🍑 (peach) out for the first weeks. In our judgment they are widely read as sexual, and they tell the other person more than you meant to.
Leave them something to answer
End a flirty message with a question that is easy to answer, or a statement they can answer without one. "Bowling rematch on Thursday? Fair fight this time, no bumpers." gives them something to push back on. Compliments that could be about anyone ("you're so pretty") give them nothing to add. A specific compliment, "you laughed at your own joke before you got to the end of it, and I liked that", does.
Questions help here for the reason Huang and colleagues give for conversations in general: follow-ups show you were listening, and our page on the first message that gets a reply applies that to openers.
Where flirting stops
If they stop matching your energy, stop raising it. Interest is easy to misread in both directions, which we cover in signs someone likes you. The lines themselves, sorted by situation, are on the page for flirty texts, along with the limits of what to send.
Frequently asked questions
How do you flirt over text without being cringe?
Which emojis are flirty?
How do you flirt with a guy over text?
How soon should you reply to a flirty text?
Sources
- Ireland, Slatcher, Eastwick, Scissors, Finkel and Pennebaker (2011), Language style matching predicts relationship initiation and stability, Psychological Science (Europe PMC record) Read October 8, 2026. Abstract only. Supports 40 speed dates with an odds ratio of 3.05 for mutual interest at higher language style matching (33.3% against 9.1% above and below the median) and 86 couples' instant messages with an odds ratio of 1.95 for relationship stability at three months (76.7% against 53.5%).
- Gesselman, Ta and Garcia (2019), Worth a thousand interpersonal words: emoji as affective signals for relationship-oriented digital communication, PLOS ONE (Europe PMC record) Read October 8, 2026. Abstract only. Supports emoji use with potential partners being associated, across two studies, with keeping a connection beyond the first date and with more romantic and sexual interactions over the previous year. The abstract describes an association, not a cause.
- Miller, Thebault-Spieker, Chang, Johnson, Terveen and Hecht (2016), Blissfully happy or ready to fight: varying interpretations of emoji, Proceedings of ICWSM, vol. 10 (AAAI) Read October 8, 2026. The AAAI page lists the paper in Vol. 10 No. 1 (2016), the Tenth International AAAI Conference on Web and Social Media; the Crossref record carries a 2021 re-index date. Supports disagreement on sentiment 25% of the time for the same rendering, more disagreement across platforms, and the potential for miscommunication. The 2017 follow-up (doi 10.1609/icwsm.v11i1.14901, abstract read) supports a controlled study of 2,482 participants finding that surrounding text did not reduce the potential for miscommunication.
- Unicode Consortium, emoji-test.txt, Emoji version 18.0 (file dated April 30, 2026) Read October 8, 2026. Supports the official names winking face, smirking face, smiling face with smiling eyes, smiling face with hearts, face blowing a kiss, see-no-evil monkey, eggplant and peach.
- Teichmann, Petrowsky, Boecker, Soliman and Loschelder (2026), How the timing of texting triggers romantic interest after the first date, Journal of Social and Personal Relationships 43(2) (Leuphana research portal) Read October 8, 2026. Abstract and record read. Supports the authors and affiliation, publication in February 2026, a pre-study of 100 to set the manipulation, the main experiment with 543 participants comparing texting immediately, the next morning or after two days, the curvilinear result with the highest relationship intentions after texting the next morning, women being more sensitive to timing, and the role of perceived reciprocity and reliability.
- Huang, Yeomans, Brooks, Minson and Gino (2017), It doesn't hurt to ask: question-asking increases liking, Journal of Personality and Social Psychology (Harvard Business School copy) Read October 8, 2026. Full text read. Supports the finding across three studies that people who ask more questions, particularly follow-up questions, are better liked, the speed dating result that a higher follow-up rate predicted agreement to a second date, people not anticipating that question-asking increases liking, and the authors' speculation that questions without reciprocal disclosure may reduce liking.