You download a dating app. You pick three photos already sitting in your camera roll, write two lines of bio, and start swiping. A few days later you have a handful of matches. One conversation opens with “hey, how's it going?”. You reply. They reply. Two messages later, nothing.

You try again the next week. Same story. A month in: thirty matches, a dozen dead threads, two people who vanished without a word — and still no date. Which is when the question shows up. Is it me, is it my photos, or do these apps just not work?

Here is what this article argues: the problem starts neither with the first message nor with the invitation. It starts well before. And more importantly, it almost always sits at one specific point in the process — not everywhere at once. Until you know which one, you are fixing things at random.

The problem is not how many matches you get

Most advice online covers one slice of the process. A photographer tells you it is all about the photos. A coach tells you it is all about the opener. Another page tells you to ask them out after exactly three days. Each of them is right about their slice, and each leaves you with the same blind spot: you still do not know where it breaks for you.

Between the moment somebody sees your profile and the moment you are sitting across from them, there are six distinct stages.

  1. Visibility — does your profile give anyone a reason to stop?
  2. The match — is there real mutual interest, or just a reflex swipe?
  3. The opener — does your first message give them something to answer?
  4. The conversation — is it progressing, or circling?
  5. The ask — is the invitation made at the right moment?
  6. Logistics — does “yes” become a date, a time and a place?

A breakdown at stage 2 will not be fixed with better messages. A breakdown at stage 6 will not be fixed with better photos. So before changing anything: at which stage do your attempts most often stop?

One thing worth knowing before we get into it, because it changes how you read the rest. These apps do work for a lot of people: a national phone survey of 1,700 Canadian adults conducted in 2018, analysed by sociologists Yue Qian and Yang Hu, found that 17.5% of coupled Canadians had met their partner online, more than two thirds of them through a dating site or app. Close to one couple in eight. So the problem is not that it does not work. It is that it works badly when you go about it badly.

Stage 0: what happens before your first swipe

There is a stage that appears on no list, and it often decides everything else: the state of mind you download the app in.

Plenty of people create an account at a particular moment. After a breakup, after a disappointment, after a month where everyone around them seemed to be paired up. Sometimes on a Sunday evening, out of boredom. Nothing shameful, nothing rare.

The problem is not the moment, it is what the moment produces. In that state, everyone does roughly the same thing: three photos already in the camera roll, a face cropped out of a group shot, two lines of bio in ten seconds, no stated intent, and straight to swiping. Each decision makes sense on its own. The trouble is that most users are doing exactly the same thing at the same time — which produces an ocean of interchangeable profiles where nobody has a reason to stop on one rather than another.

An impulsive sign-up also tends to come with a fuzzy intention, and that costs more. If you do not know yourself whether you want something serious, something casual or just a distraction, you cannot express it — and you will spend weeks talking to people whose expectations have nothing to do with yours. That is not a flirting failure, it is a routing problem.

Stage 1: your profile is your first conversation

This is the most important point here. Until your profile gives someone a reason to act, none of the rest can happen. No like, no match, no conversation, no date. Your profile is not paperwork before the real game — it is the first message you send, and you send it to everybody at once.

A good profile is not a photogenic person's profile, it is a legible one. Somebody scanning it for three seconds should be able to answer four questions without effort: what does this person look like, what does their life look like, what seems to matter to them, and would I have anything to say to them.

  • A sharp, recent main photo with your face visible. Not a questionable crop out of a group shot, not sunglasses in all five photos. This is not about beauty, it is about uncertainty: a face you cannot see is a doubt nobody resolves.
  • Two or three photos with context. A place, an activity, something you care about. They are not there to impress, they are there to give people something to grab: a photo of you climbing or cooking offers ten times more openings than a flawless bathroom selfie.
  • A short but specific bio. “I like travelling, good food and laughing” says nothing — eight people out of ten could write it. One concrete detail does the job.
  • A clear intention. Saying what you want does not scare off the right people. It scares off the wrong ones.

A specific profile often gets fewer likes than a consensus one, and that is good news. Fifty matches with people who have nothing to say to you are worth less than three matches with people who read your bio and decided it interested them.

What suspiciously polished profiles teach us anyway

You have probably come across profiles that seem to attract attention effortlessly: consistent photos, careful lighting, everything understandable in three seconds. Some are genuine. Others are fake, built precisely to maximise attention — that is a reality of these platforms.

The point is not the lazy conclusion that “the profiles that work are fake”. The lesson is elsewhere: those profiles have one thing in common, they were built to be understood instantly. Nothing is left to chance; every element removes a doubt. If people acting in bad faith invest that much energy in presentation, it is because presentation works. Someone genuine would be wrong to skip it. The difference is not the care taken — it is that your profile tells the truth.

Stage 2: liking everyone does not create more connections

When nothing happens, the instinctive response is to raise the volume. Like more, faster, without reading. It is logical enough: if one profile in twenty replies, see two hundred. Except it backfires.

You stop reading. After a few dozen swipes the gesture goes automatic and you only see the first photo. You skip people who would have suited you, and you like profiles you know nothing about.

You get empty matches. A match born from a reflex swipe on both sides carries no information — nobody knows why they liked the other. That is exactly the soil “hey, how's it going?” grows in, followed by silence.

Abundance degrades judgment. A lab experiment with 152 American undergraduates by Jonathan D'Angelo and Catalina Toma found that people choosing from twenty-four profiles were less satisfied with their choice than those choosing from six, and more likely to reverse it a week later. Fictional profiles, laboratory setting: a demonstration of a mechanism, not a law. But anyone who has spent an evening swiping will recognise the mechanism.

Worth a nuance that cuts against the usual talking point: asked in July 2022 by the Pew Research Center, only 37% of American users felt they had “too many” options, against 43% who felt they had about the right number. Overload becomes a problem when it happens to you, not when you choose it.

Stages 3 and 4: why conversations die after three messages

This is where the largest number of people drop off, and where the stories we tell ourselves are the least fair.

One data point puts things in perspective. An analysis of nearly two million conversations on a mobile dating app by Jennie Zhang and Taha Yasseri of the Oxford Internet Institute found that roughly half of first messages get a reply. It is a preprint based on American data from 2013-2015 on an unnamed app, so take it as an order of magnitude. But the order of magnitude says the important part: one message in two going unanswered is not a signal about you, it is how the medium works.

That said, when every conversation dies in the same place, there is usually an identifiable cause.

The inert opener. “Hey”, “How's it going?”, “Nice photo” give the other person nothing to answer. They would have to do all the work to get things moving — and they will not, because they have eight other threads in the same state.

The interview. Stacking “what do you do?”, “where do you live?”, “any siblings?” without ever offering anything of your own turns a conversation into a form.

The imbalance. When one person writes paragraphs and the other answers “haha”, it is already over, even if it drags on a few more days.

The rhythm. Replying within thirty seconds for six straight hours exhausts a thread; replying every four days chills it.

Mismatched intentions. Sometimes the conversation is good, funny, easy — and it dies anyway, because one of you wants a relationship and the other does not. That is not a technique problem.

And ghosting

Its real frequency is uncertain. A literature review published in 2024 by researchers Gili Freedman and Darcey Powell reports rates ranging from 23% to 90% of people having been ghosted, depending on the study — a spread explained mostly by who was surveyed. In the general American population, a YouGov poll of 7,180 adults in 2021 found 19% said they had been ghosted by someone they were dating; among active app users the share climbs much higher. So: yes, common; no, not universal.

The reasons are rarely the ones we imagine: lost interest, but also parallel conversations, discomfort at delivering a rejection, plain fatigue. And sometimes something more serious — an experimental study by Gili Freedman and colleagues in 2022 found that safety concerns markedly increase the intention to disappear rather than reject someone directly. Those are hypothetical scenarios, not observed behaviour, but the idea deserves airtime: silence is not always contempt, sometimes it is an exit.

Stage 5: an opener that can actually lead somewhere

There is no magic line, and be wary of anyone selling you one. There is, however, a method that clearly improves your odds, and it takes five moves.

  1. Actually read the profile. Thirty seconds. Look for the specific thing — not “she likes travelling”, but “she has a photo from a festival I know”, “he has been learning piano for six months”.
  2. Hook onto that thing with an open question — one that cannot be answered yes or no.
  3. Offer something of your own. One sentence is enough. It turns a question into an exchange.
  4. Build on the answer. This is the step almost everyone skips. If they tell you they took up piano during a burnout, the conversation is right there — not in your next prepared question.
  5. Remember what was said. The cat's name, the hometown, the project. Coming back three days later to a detail mentioned in passing is the most effective interest signal there is, and it costs nothing.

Two examples, deliberately ordinary — a good opener does not need to be clever, it needs to be addressed to one particular person.

“That photo on Mont-Royal in the middle of winter honestly impressed me. I'm more of a disappear-from-November-to-March type. Is that a regular thing or a one-off act of madness?”

“We have the same Vietnamese place in our photos, so I'll ask the only question that matters: do you always order the same thing or do you switch it up? Mine hasn't changed in four years.”

Stage 6: when — and how — to ask

There is no “ask after three days” rule. Numbered rules are comforting, but they ignore the only thing that matters: is the interest mutual, and is there enough trust to meet. What can be said is that you should neither ask immediately nor wait indefinitely. Asking within the first two messages skips trust, and plenty of people — with good reason — will pass. Talking for three weeks without ever asking turns the thread into a pen-pal arrangement: pleasant, comfortable, going nowhere.

The signs it is time: replies get longer; questions flow both ways; personal things surface unprompted; a shared sense of humour appears; a place or an activity keeps coming up in the conversation. When three of those five are there, ask.

One thing to phrase carefully. Someone who struggles to keep a written conversation alive is not necessarily someone who struggles in person — plenty of people are far better face to face, and that is entirely legitimate. But if none of your conversations ever take off while you already have direct access to the person, do not assume meeting in person will automatically fix it. Online silence usually points to one of five things: weak interest, wrong rhythm, mismatched expectations, a generic conversation, or not listening. None of the five disappears by changing the setting.

When you do ask, be concrete, leave an exit, and do not request a disproportionate commitment:

“I've enjoyed this. Would you be up for a coffee one evening this week? I'm free Thursday or Saturday, but tell me what works for you.”

The stage nobody writes about: turning “yes” into a confirmed date

This is where a large share of dates die, and almost no article mentions it. “Yes, I'd love to!” is not a date. A date is a day, a time and a place. Until those three exist, you have nothing — and “sure, sometime” is the most common polite way of avoiding a no.

Lock it in within twenty-four hours. When they say yes, immediately offer two concrete slots and a specific place: “Great — Thursday 6pm or Saturday 2pm, at the café on Saint-Denis?” Whatever comes back will be clear.

Do not leave four days of silence between the yes and the date. That is the window where momentum dies and doubt sets in. One light message the day before — not three — keeps the thread alive without applying pressure.

If they reschedule, offer an alternative right away. A reschedule is not a refusal. A reschedule nobody follows up on becomes one.

The date should extend the conversation, not replace it

The venue should not come from a generic list, but from what you have already told each other — that is what separates a first date from a job interview. If they mentioned an obsession with a particular kind of coffee, the place picks itself. If you spent twenty messages on exhibitions, the museum is not an idea, it is a conclusion.

Two principles hold: short and public. A coffee, a walk somewhere busy, an early drink, an exhibition, a market. One or two hours is plenty. A three-hour dinner puts unnecessary pressure on two people who do not know each other, and makes leaving awkward if it does not click.

In Montreal and across Canada, two realities are worth planning around rather than suffering. Winter runs five months — suggesting a walk on the Plateau in February is not romantic, it is an endurance test; a café, a bookstore, a museum or an early drink work better for half the year. And geography matters: when one of you is in Laval and the other on the South Shore, a badly chosen spot turns a one-hour coffee into a three-hour expedition. Meeting halfway, near transit, seriously improves the odds that the date happens at all — and that there is a second one.

Safety is part of a good first date

This is not a moralising add-on at the end: it is a condition for the meeting going well, for both people.

The subject is not theoretical. An SSRS survey of 2,012 American adults in January 2026 found that 43% think it is not very or not at all safe to meet someone in person you met online, with a sharp split: 55% of women against 30% of men. And among women under 50 who have used a dating app, the Pew Research Center reported in 2022 that 56% had received unsolicited sexually explicit messages or images. Those figures are American and the Canadian context is not identical, but they explain something better understood than worked around: if someone takes precautions before meeting you, it is not suspicion of you. It is ordinary practice, and the best response is to make it easy.

In practice, on both sides: a public, busy place at a reasonable hour; your own transportation both ways; telling a trusted person where you are going, with whom, and when you expect to be back; keeping your home address for later; respecting a no or a hesitation without pushing; leaving if you are uncomfortable, without justifying it; and reporting and blocking problematic behaviour, which also helps other people.

How Kiro tries to make this path feel more natural

Everything above works on any app. But each of those stages asks for effort the app does not help you with: it shows you profiles, counts your matches, and leaves you to work out the rest.

That is where we started. Not from the idea that we could predict love better — nobody can do that — but from the idea that before helping someone find a person, you have to help them understand what they are looking for and present themselves better.

Onboarding that asks real questions. Instead of letting you build an empty profile in forty seconds, Kiro walks you through a limited set of useful questions: what you are looking for right now, your expectations, your priorities, what draws you to someone, the settings you feel at ease in. The point is not to lengthen sign-up, it is to avoid stage 0.

A test based on the Big Five model. The five-factor model is the most established framework in personality psychology: five relatively stable dimensions — openness, conscientiousness, extraversion, agreeableness, emotional stability. To be clear about what it does and does not do: it guarantees no romantic compatibility, and nothing lets anyone predict a relationship from five scores. What it offers is more modest — Kiro does not only try to learn who you find attractive, but also what you value in a relationship, so suggestions are less random.

An assisted bio. Many people rush their bio not out of laziness but because they do not know what to write. Kiro drafts a first version from your own answers — not “an AI writes it for you”, but a coherent starting point that solves the blank page, which you then edit until it sounds like you.

Places that look like you. You can attach to your profile the kinds of places where meeting people feels natural to you: a café, a park, a library, a gym, a cultural event, a community, a church. It applies to romantic, friendly and community connections alike. And to remove any doubt: only the name of the place is visible — never a private address, never a pin on a map, and residential addresses are automatically rejected.

An affinity indicator. Kiro shows a marker reflecting how much you have actually learned about each other over the course of the exchange. It is not a scientific measure of love and should not be read as a verdict — it is a visual nudge toward exactly what stage 5 describes: ask questions, listen to the answers, remember what was said.

Games instead of “hey, how's it going?”. This is the part most directly tied to the dead-conversation chapter. Rather than leaving you alone in front of an empty text field, Kiro offers conversational experiences: paired questions that surface what you have in common, a reworked truth-or-dare to get started without effort, a prompt game for when a thread stalls. One of them, before you meet, checks what you have actually retained about each other — a colour, a place, a detail dropped mid-conversation. The goal is not to turn a relationship into an exam, but to reduce the awkwardness of the first message and to reveal whether both people are listening. When the result is good, moving to a real date becomes obvious — and that is where Kiro offers to plan it rather than letting you circle.

None of this guarantees a date. But each piece targets a specific stage where people drop off today.

The short version

  • Your profile is your first conversation. Until it gives someone a reason to act, nothing else can happen.
  • Swipe count is not the metric. Ten profiles read beat a hundred skimmed; volume manufactures empty matches.
  • A conversation is an exchange, not a form. Open questions, something of your own, build on the answers, remember what was said.
  • Interest has to be mutual. If you are carrying all the effort, you already have your answer.
  • The right moment shows up as signals, not on a clock.
  • “Yes” is not a date. A day, a time, a place — otherwise you have nothing.
  • The venue should come from your conversation, not a list. Short, public, easy to reach.
  • Safety is part of a good date, for both people.
  • Ghosting is almost never a verdict on you.

Where that leaves you

There is nothing abnormal about it not working right away. The data says so plainly: asked by the Pew Research Center in October 2019, Americans who had used a dating app in the previous year came away more often frustrated (45%) than hopeful (28%). And yet in July 2022, 53% of people who had ever used one rated the experience positive overall. Both are true at once. It is frustrating, and it works anyway.

What actually changes your results is neither hours spent swiping nor luck. It is where you put the effort. A profile that says something true, likes you chose rather than reflexed, conversations where you listen as much as you talk, and an invitation made when the interest is visibly shared: none of it guarantees anything, but it changes the odds completely.

You may not need to swipe more. You may need an app that helps you present yourself better and build exchanges that go somewhere. That is what we are trying to build with Kiro, and the beta is open if you want to see what it looks like.

Experiences on dating apps vary by person, intention, region and platform. The advice here aims to improve the quality of a profile and of conversations; it guarantees no particular number of matches or dates. This article is not psychological or therapeutic advice.

Sources: Qian & Hu, Journal of Marriage and Family, 2025 (2018 data) · Pew Research Center, 2020 · Pew Research Center, 2023 · SSRS, 2026 · YouGov, 2021 · Freedman & Powell, Social and Personality Psychology Compass, 2024 · Freedman et al., Journal of Experimental Social Psychology, 2022 · D'Angelo & Toma, Media Psychology, 2017 · Zhang & Yasseri, preprint, 2016