Why People Ghost, and What to Send Instead

I spend most of my working life thinking about messages that never got sent.

That’s the job. I run a writing app, and the feedback that sticks with me is almost never about grammar. It’s someone describing a text they drafted eleven times and then deleted. A reply they owed a friend three weeks ago and now feel too embarrassed to send. The “hey, I don’t think this is working” that turned into nothing at all.

Which is how I ended up reading forty-odd papers on ghosting over the past few weeks. I expected the research to confirm the obvious story: people ghost because they don’t care, and the internet made them worse. Some of that holds up. A lot of it doesn’t. And the part that surprised me most is directly useful if you’re the one holding the phone right now, trying to figure out what to type.

Here’s what I found.

What ghosting actually means, and what it doesn’t

Researchers are stricter about this word than we are. In the literature, ghosting means unilaterally ending a relationship by cutting off all contact¹, without explanation. Gili Freedman and Darcey Powell, who between them have produced most of the good work on this², note that definitions differ on whether it has to be sudden or can be a slow fade, but they all share that core: communication stops, and nobody says why.

The useful bit for anyone spiralling right now: matching with someone and never hearing back is probably not ghosting. Freedman and Powell are explicit that some back and forth has to happen first. One unreturned opener on a dating app is a non-event. Your brain will tell you otherwise. Your brain is wrong.

The word itself is newer than the behavior. Merriam-Webster added the verb sense in 2017, and search interest peaked around 2019. But avoidance and withdrawal were catalogued as breakup strategies by Leslie Baxter back in 1982³, long before anyone had a smartphone to hide behind. So when people say ghosting is a symptom of modern dating, I’d push back a little. What’s new is the name, the visibility, and the read receipt.

How common is ghosting, actually

This is where I got annoyed at most of the articles I read.

You’ll see stats like “85% of dating app users have been ghosted” thrown around constantly. I went looking for the source. I couldn’t verify it. What I could verify is that Freedman and Powell, reviewing every study that reported a rate, found the range for having ghosted someone runs from 18.9% all the way to 92.0%, and being ghosted from 23.0% to 89.7%². That’s not a statistic. That’s a shrug with decimal places.

The gap comes down to who you ask. Recruit from dating apps and you get enormous numbers. Recruit from the general adult population and it drops hard. In Freedman’s 2019 general-population studies, 21.7% had ghosted a romantic partner and 25.3% had been ghosted¹.

The widely quoted Forbes Health figure⁴, that 76% of people have ghosted or been ghosted, comes from an OnePoll survey of 5,000 US adults who dated in the past five years. Big sample, real finding, but it’s a brand-commissioned poll of people who were actively dating, not a random slice of humanity. Worth citing. Not worth treating as gospel.

My honest read: ghosting is common, it’s more common in early and casual connections, and anyone selling you a precise number is guessing.

Why people ghost: the finding that changed my mind

This one I’ve reread three times.

In 2024, Yina Park and Nadav Klein published eight experiments⁵ plus three pilot studies in the Journal of Experimental Psychology: General. They looked at whether ghosters actually care about the people they ghost. The answer, consistently, was yes. More than the ghostee assumes. It held when people recalled past ghosting, when they ghosted in real time in the lab, and even when refusing to ghost cost them money.

Their conclusion is worth sitting with: ghosting is social rejection without explanation or feedback, but not without care. People who go quiet often do it partly as a clumsy attempt to avoid hurting you, and the person on the receiving end reads the silence as proof of total indifference.

Both parties end up with the wrong picture of each other. The ghoster thinks silence is the gentle option. The ghostee reads it as the cruellest one.

Which connects to the other finding I keep coming back to. Freedman and Powell tie ghosting directly to what Roy Baumeister called scriptlessness: people don’t know what to say when they have to reject someone². There’s no script. Nobody teaches you the words. And rejection is one of the few social situations where being articulate genuinely matters and most of us are winging it.

I’ll be upfront that this is convenient for a guy who sells a writing app. It’s also just what the data says. A large share of ghosting looks less like coldness and more like a wording problem that grew teeth.

Why being ghosted hurts more than a clean rejection

Three separate research groups, using three different methods, reached the same conclusion here. That kind of convergence is rare enough that I trust it.

Kipling Williams’ work on ostracism⁶ gives us the framework. Being ignored threatens four things at once: your sense of belonging, your self-esteem, your sense of control, and your feeling that you matter. Not one of those. All four.

Christina Leckfor and colleagues at the University of Georgia ran three preregistered studies⁷ and found that being ghosted left people with significantly lower satisfaction of those needs than being rejected directly. They also found that a high personal need for closure didn’t stop anyone from ghosting others, but it dramatically amplified the pain of being on the receiving end.

Marco Pancani’s team in Italy compared ghosting, orbiting, and direct rejection⁸, and found belonging and control took a bigger hit from ghosting than from being told no. Then in 2025, Telari, Pancani and Riva ran a multi-day diary study⁹ rather than asking people to remember old wounds. Their finding: both hurt, but ghosting’s effects last longer, because the uncertainty blocks the coping process from starting at all.

Pauline Boss has a name for this that I find clarifying. Ambiguous loss. A loss you can’t confirm, so you can’t grieve it. LeFebvre and colleagues applied it directly to ghosting¹⁰, describing the strange condition of someone being physically and psychologically gone while still fully present on your phone.

That’s the specific cruelty. Not the rejection. The fact that you can’t even be sure a rejection happened.

One nuance before anyone catastrophizes. A 2026 randomized controlled trial from Baylor¹¹ took 112 emerging adults, had them text a lab confederate for two hours, then randomly assigned some of them to be ghosted, some to get a polite ending, and some to nothing. Objective sleep data from Oura rings showed no meaningful difference. Two hours of texting with a stranger who then disappears does not wreck your sleep. The damage scales with what you’d invested. Which sounds obvious, and is still nice to see measured.

Friendship ghosting is more common than the dating kind

This part barely gets written about, and it surprised me most.

In Freedman’s 2019 data, friendship ghosting was more common than romantic ghosting in the same sample. 31.7% had ghosted a friend versus 18.9% a romantic partner. 38.6% had been ghosted by a friend versus 23.0% by a partner¹.

More interesting, the reasons are different. Michaela Forrai’s team at the University of Vienna ran a two-wave panel study¹² with 978 young people and found that romantic ghosting was predicted by communication overload, that swamped feeling of too many chats. Friendship ghosting wasn’t. Friendship ghosting was predicted by the ghoster’s own self-esteem and withdrawal.

And there’s a sting in the tail. People who reported ghosting friends showed increased depressive tendencies four months later. Ghosting romantic partners showed no such effect. As Forrai put it, people who said they had ghosted friends were more likely to report increased depressive tendencies at follow-up.

I don’t want to overstate one panel study. But it lines up with something a lot of people recognise: quietly dropping a friend because you’re exhausted and behind on replies tends to make you feel worse, not lighter.

Ghosting at work, in both directions

Ghosting has fully colonised hiring, and the numbers are wild in both directions. Indeed’s 2023 survey of thousands of job seekers and employers¹³ found 78% of candidates had ghosted a prospective employer, up from 68% the year before, and 40% said an employer had ghosted them after a second or third interview. Then there’s career catfishing, where someone accepts a job and simply doesn’t show up. A CV Genius survey of 1,000 UK employees¹⁴ put that at 34% of Gen Z, 24% of millennials, 11% of Gen X and 7% of boomers.

Before anyone writes a column about kids these days, look at the other side. ResumeBuilder surveyed 1,641 hiring managers in May 2024¹⁵. 40% said their company had posted a fake job listing in the past year, and three in ten had one live at that moment, for reasons including projecting growth and making existing employees feel replaceable. Greenhouse has estimated 18% to 22% of online job ads are fake or unfilled.

Lawmakers noticed. The FTC formed a Joint Labor Task Force in February 2025 with deceptive job advertising as a priority, several US states have introduced ghost job disclosure bills, and Ontario has enacted one taking effect in 2026. Most of the American bills are still in committee rather than signed¹⁶, which a lot of coverage gets wrong.

Does every culture ghost the same way

If you’ve ever wondered whether your culture ghosts more than others, there’s finally real data. It landed five days before I wrote this.

Freedman and Powell published the first systematic cross-national comparison¹⁷, covering twelve countries across two studies. In the first, 885 young adults from Canada, Kenya, Mexico, Nigeria, South Africa and the US. Around 96% recognised the behavior, though the vocabulary shifts by region. In South Africa, some participants used “mizing.”

Acceptability varied a lot. Kenyan, Nigerian and South African participants found it more acceptable than American, Canadian and Mexican ones. Everyone agreed on one thing: short relationships are fair game. Across all six countries the estimated likelihood of ghosting a short-term connection averaged 52%, versus 20% for a long-term one.

The second study, 1,741 adults across Brazil, Ghana, India, Indonesia, Kenya, the Philippines and Ukraine, threw up something I love. Attitudes and behavior don’t track. Indian participants reported some of the highest acceptance of ghosting and the lowest actual rate, roughly 16% for romantic partners. Brazilians reported low acceptance and low intentions, and yet about 53% had ghosted a friend.

So the tidy theory, that direct cultures ghost less and indirect cultures ghost more, doesn’t survive contact with the data. What people think about ghosting and what they do about it are two different questions.

One more, since we work in Turkish and I went looking. Betül Kabasakal and Elif Cimsir ran a vignette experiment with 224 Turkish adults¹⁸ and found ghosting was judged equally inappropriate in friendships and romantic relationships. There’s also now a validated Turkish version of the Ghosting Questionnaire¹⁹, so the research base there is finally catching up.

When ghosting someone is the right call

I want to be careful here, because “always send the message” is bad advice delivered universally.

Freedman, Hales, Powell, Le and Williams ran a registered report²⁰ specifically on how the target’s gender and perceived danger shape rejection decisions. The broader picture is that women ghost more often in part because a direct romantic rejection can trigger retaliation. If someone has been aggressive, if something feels off, if you have any reason to think a clear no will escalate, disappearing is a legitimate self-protective choice and nobody should make you feel bad about it.

There’s also a reasonable argument that low-investment interactions don’t require a formal exit. Some participants in Timmermans’ dating app research²¹ argued exactly that, and that a laboured rejection message can land worse than silence. I don’t fully agree, but it isn’t a strawman position.

Safety first, always. Everything below is for the ordinary case where you just don’t know what to write.

What to send instead of ghosting

This is the part I actually wanted to write, so let me be concrete.

Skip the apology. This is the single most counterintuitive finding in the rejection literature and I’ve changed how I write these messages because of it. Across three sets of studies, Freedman and colleagues found that apologies increased hurt feelings²² and increased the recipient’s sense that they had to express forgiveness, without actually increasing forgiveness. As Freedman put it, people apologise with good intentions, and it makes the other person feel worse and obliged to forgive them before they’re ready. So “I’m so sorry, but” is doing damage. Cut it.

Say the thing early. Burying the point in paragraph three so it lands softly does the opposite. The reader is scanning for the verdict and every sentence before it is noise.

Use real positive regard, not fake hope. The Responsive Theory of Social Exclusion, developed by Freedman, Williams and Beer²³, holds that explicit rejection with genuine warmth protects the recipient’s needs better than ambiguity, and costs the sender less emotionally too. “I enjoyed meeting you” is fine if it’s true. “Maybe down the line” is not, if it isn’t.

Give it a few more words than feels comfortable. Curt reads as contempt. You don’t need a paragraph, but a single clipped line lands like a door slamming.

Don’t wait for the perfect version. Day four is much harder than day one, and the awkwardness compounds. This is the trap, and it’s the reason ghosting looks like a character flaw when it usually starts as a scheduling problem.

Here’s the whole thing, in the length it should be:

Hey, I’ve really enjoyed talking to you, but I don’t think this is going anywhere romantic for me. Wanted to say it properly rather than just go quiet. Wishing you well.

That took eleven seconds. Nobody is destroyed by it. And there’s evidence you can learn this: Okland, Freedman and Beer showed that a roughly twelve-minute lesson²⁴ measurably improved people’s ability to write better rejections. It’s a skill, not a personality trait.

The one nobody writes about: you already went quiet

Every article on this assumes you’re standing at the fork. Most people reading are past it. You went quiet nine days ago, or three weeks ago, and now the silence itself is the embarrassing thing.

The instinct is to explain the gap, which is exactly what keeps you from sending anything, because there’s no good explanation for three weeks. So don’t build one. Name it in four words and move on:

Hey, I disappeared and that was a rubbish thing to do. I should have said earlier that I wasn’t feeling it romantically. Sorry for leaving you hanging.

Note that this one keeps the apology, and that isn’t a contradiction. Freedman’s finding is about apologising for the rejection itself, which pushes the other person to forgive you for something that isn’t a wrong. Apologising for the silence is different. The silence genuinely was a wrong, and it’s the one thing they’re owed.

They may not reply. That’s fine. The message isn’t a negotiation.

If it’s a friend rather than a date

Harder, because there’s no equivalent script at all. Nobody has a template for “we drifted and I let it happen.” The move that works is dropping the audit and going straight to the present:

I’ve been a terrible friend at replying and I’m sorry. Nothing happened, I just fell off. I’d like to fix it if you’re up for it. Coffee this month?

Explaining why you went quiet is almost always less useful than showing you’d like to stop.

If you’re the one who got ghosted

Most advice on this is either “closure is essential” or “just move on,” and the research supports neither cleanly.

Leckfor’s team found that a high need for closure sharply amplifies the pain of being ghosted⁷. That’s real. But the closure you’re waiting for is being withheld by someone who has already shown you they can’t have this conversation, so waiting is a bad bet. Pauline Boss’s whole point about ambiguous loss is that the resolution has to be constructed rather than received¹⁰.

Which is where the Park and Klein finding actually earns its keep⁵. Your ghoster probably did care. Not enough to write to you, but the story you’re running, that you meant nothing, is measurably wrong more often than it’s right. That isn’t comfort exactly. It’s just a more accurate account than the one your brain is building at 1am.

Two practical things. Send one message if you want to, not four: something like “I’m guessing this has run its course, no hard feelings, but let me know either way” gives them an exit and gives you a line under it. Then stop. And if they resurface weeks later as though nothing happened, you’re allowed to not pick it back up.

The other reframe worth keeping: silence tells you about their capacity for hard conversations. It’s a fact about them. Whatever your brain does with it at 1am, it is not a verdict on you.

What I actually think

Ghosting gets framed as a moral failure, and after reading all of this I think that’s mostly wrong. Cruelty exists, and some people ghost because they don’t care. But the bulk of it looks like ordinary people with no script, no practice, and a rising sense of dread, choosing the option that requires no words.

The gap between “I don’t know how to say this” and “I’m going to say nothing” is smaller than it should be. Closing that gap is roughly the entire reason Subtext exists, so consider that disclosed.

If you’re sitting on a message you owe someone, the research is unusually clear about the trade. Eleven seconds of your discomfort, or weeks of theirs.

Send the eleven seconds.

Think I’ve let ghosters off too easily? Tell me on LinkedIn.

Samet Durgun is the co-founder of Subtext, an app that catches the emotional tone of your messages and rewrites them in your own voice. He’s based in Berlin.

Sources

Every link above goes to the primary source where one exists.

Freedman, G., Powell, D. N., Le, B., & Williams, K. D. (2019). Ghosting and destiny: Implicit theories of relationships predict beliefs about ghosting. Journal of Social and Personal Relationships, 36(3), 905-924. doi.org/10.1177/0265407517748791
Freedman, G., & Powell, D. N. (2024). Ghosting: A common but unpopular rejection strategy. Social and Personality Psychology Compass, 18(12). doi.org/10.1111/spc3.70026 · Full text PDF
Baxter, L. A. (1982). Strategies for ending relationships: Two studies. Western Journal of Speech Communication, 46(3), 223-241. doi.org/10.1080/10570318209374082
Forbes Health / OnePoll survey of 5,000 US adults, 2023. Survey Reveals ‘Ghosting’ Impacts 76% Of People Who Are Dating
Park, Y., & Klein, N. (2024). Ghosting: Social rejection without explanation, but not without care. Journal of Experimental Psychology: General, 153(7), 1765-1789. doi.org/10.1037/xge0001590
Williams, K. D., & Nida, S. A. (2011). Ostracism: Consequences and coping. Current Directions in Psychological Science. doi.org/10.1177/0963721411402480
Leckfor, C. M., Wood, N. R., Slatcher, R. B., & Hales, A. H. (2023). From close to ghost: Examining the relationship between the need for closure, intentions to ghost, and reactions to being ghosted. Journal of Social and Personal Relationships, 40(8), 2422-2444. doi.org/10.1177/02654075221149955 · University of Georgia summary
Pancani, L., Aureli, N., & Riva, P. (2022). Relationship dissolution strategies: Comparing the psychological consequences of ghosting, orbiting, and rejection. Cyberpsychology, 16(2). doi.org/10.5817/CP2022-2-9
Telari, A., Pancani, L., & Riva, P. (2025). Ghosting and direct rejection over time. Computers in Human Behavior, 172. doi.org/10.1016/j.chb.2025.108756
LeFebvre, L. E., Rasner, R. D., & Allen, M. (2020). “I guess I’ll never know…”: Menendez, ambiguous loss, and ghosting. Personal Relationships, 27(2). doi.org/10.1111/pere.12322 · Ambiguous loss theory: Boss, P. (1999), University of Minnesota.
Langlais, M. R., Pons, A., Dechert, O., Murvich, A. M., & Bigalke, J. A. (2026). Resilience to ghosting? A randomized controlled trial examining the consequences of ghosting on sleep quality in potential romantic relationships. Frontiers in Psychology. doi.org/10.3389/fpsyg.2025.1742356
Forrai, M., Koban, K., & Matthes, J. (2023). Short-sighted ghosts: Psychological antecedents and consequences of ghosting others within emerging adults’ romantic relationships and friendships. Telematics and Informatics, 80. doi.org/10.1016/j.tele.2023.101969 · PsyPost summary
Indeed survey of job seekers and employers, conducted 2023, reported by CNBC (February 2024). Ghosting gets more common in the job market
CV Genius survey of 1,000 UK employees, reported by Fortune (January 2025). Welcome to career catfishing
ResumeBuilder.com survey of 1,641 hiring managers, May 2024. 3 in 10 companies currently have fake job postings listed · Greenhouse estimate reported by CNBC
Congressional Research Service, “Ghost” Job Postings · Forbes: Ghost job ads are latest employer tactic targeted by state lawmakers
Freedman, G., & Powell, D. N. (2026). A cross-national comparison of ghosting knowledge, attitudes, intentions, and behaviors in friendships and romantic relationships spanning twelve countries. Journal of Social and Personal Relationships. doi.org/10.1177/02654075261463579 · PsyPost summary
Kabasakal, B., & Cimsir, E. (2025). Ghosting perceptions across gender, relationship contexts, and prior experiences: Insights from a vignette study. Journal of Social and Personal Relationships, 42(10), 3101-3126. doi.org/10.1177/02654075251360612
Atalar, A. E., Çolakoğlu, N., Bağişlioğlu, Z., Ammar, A., & Jahrami, H. (2025). Turkish adaptation of Ghosting Questionnaire: Validity and reliability study. BMC Psychology, 13, 397. doi.org/10.1186/s40359-025-02677-1
Freedman, G., Hales, A. H., Powell, D. N., Le, B., & Williams, K. D. (2022). The role of gender and safety concerns in romantic rejection decisions. Journal of Experimental Social Psychology, 102. doi.org/10.1016/j.jesp.2022.104368
Timmermans, E., Hermans, A. M., & Opree, S. J. (2021). Gone with the wind: Exploring mobile daters’ ghosting experiences. Journal of Social and Personal Relationships, 38(2), 783-801. doi.org/10.1177/0265407520970287
Freedman, G., Burgoon, E. M., Ferrell, J. D., Pennebaker, J. W., & Beer, J. S. (2017). When saying sorry may not help: The impact of apologies on social rejections. Frontiers in Psychology, 8, 1375. doi.org/10.3389/fpsyg.2017.01375 · Freedman quote via ScienceDaily
Freedman, G., Williams, K. D., & Beer, J. S. (2016). Softening the blow of social exclusion for both targets and sources: The Responsive Theory of Social Exclusion. Frontiers in Psychology, 7, 1570. doi.org/10.3389/fpsyg.2016.01570
Okland, S., Freedman, G., & Beer, J. S. (2026). Knowing versus doing: How much can training help people soften the blow of social rejection? Personality and Social Psychology Bulletin. doi.org/10.1177/01461672261439420 · Authors’ plain-English write-up at SPSP

Freedman, G., & Powell, D. N. (2024) also served as the primary review source for prevalence ranges and definitional debates throughout this piece. Emotional experiences of ghosters and ghostees drawn from Freedman, Powell, Le & Williams (2022), Journal of Social Psychology, doi.org/10.1080/00224545.2022.2081528. Breadcrumbing prevalence from Navarro et al. (2020), PMC7037474.