If the tools work, why do so many women describe 3 years of steady use and nothing to show for it? The complaint is consistent enough across ages and cities that it is not a story about individual bad luck. Several features of these platforms work against the outcome their users say they want, and none of those features is hidden.
The Ratio Problem
Most general platforms have more men than women, usually somewhere between 55% and 75% male depending on the service and the city. That imbalance sounds as though it should favour women, and in raw arithmetic it does. A woman receives far more approaches than a man does.
What it produces in practice is a different market. Men, facing low response rates, send high volumes with low effort. Women, facing high volumes, screen hard and fast. Both sides then adapt to the other’s adaptation, and the result is a system where a thoughtful message from a serious person is statistically indistinguishable from 40 copy-pasted ones. The imbalance varies sharply by city and by age band. A woman of 45 in a mid-sized town is working with a very different pool from a woman of 28 in a capital, and the platforms present both with the same interface and the same promises.
Message Volume and Intent
Volume is the part that gets misread as abundance. Survey work on app users repeatedly finds a majority of women reporting that the quantity of contact is overwhelming, against roughly half that share of men. The gap holds across countries and across age bands.
An inbox of 200 messages contains no more serious prospects than an inbox of 20 if the composition is the same. It simply takes longer to process, and the processing itself is unpaid work. The reason women give for leaving is usually the time it takes.
The Limits of a Profile
Screening at speed means screening on proxies, and the proxies are weak. A woman filtering for a high-value man is usually filtering for evidence of stability and follow-through, and neither of those qualities appears anywhere in a profile. What appears is a job title, a height, 6 photographs and a sentence about travel.
The proxies people fall back on are the ones a screen can carry. Height is one of them. Reliability only shows itself across months of ordinary behaviour. That mismatch does more damage to a search for something serious than any algorithm does, because it filters on exactly the traits that predict a good first impression and almost nothing that predicts a good third year.
The Rejection Mindset
Tila Pronk and Jaap Denissen at Tilburg University ran 3 studies on what continuous swiping does to a person’s standards. They found acceptance rates fell steadily as participants worked through more profiles, with an average drop of about 27% between the first options seen and the last, alongside rising pessimism about being accepted themselves.
They named the pattern a rejection mindset, and it was stronger in women than in men. The longer a woman searches on a platform built around volume, the less likely she becomes to accept anybody at all, independent of who is actually in front of her. The mechanism is dissatisfaction with the images she has already seen and a growing expectation of being turned down herself, both of which feed the next decision.
The evidence points both ways. A more recent study in the Journal of Social and Personal Relationships found that a larger pool increased participants’ desire to pursue a chosen match, with no sign that seeing more profiles made people pickier. The effect probably depends on how a given platform presents its options and on how quickly it asks for a decision.
Safety Costs and Withdrawal
A 2022 Australian Institute of Criminology survey found abuse on the apps reported by 3 in 4 users, and the pattern was heavily gendered. Around 38% of women aged 18 and 19 reported gender-based harassment, and 1 in 3 reported unwanted sexually explicit material.
The effect on serious relationship formation is indirect and large. Women who have been harassed screen harder, meet in person later, and disengage sooner. Every one of those responses is rational, and every one of them lowers the odds of the slow, low-stakes acquaintance that most durable relationships start from. Caution that would be excessive in a workplace or a walking group is the correct setting on a platform where a quarter of approaches are hostile, and there is no way to apply it selectively.
Displacement and Population-Level Effects
The platforms did not arrive alongside the old ways of meeting people. Michael Rosenfeld’s work on how couples meet shows that online introductions overtook meeting through friends around 2013 and then kept climbing, while the older routes fell away.
That matters because the old routes had a screening step built into them. A friend who introduces 2 people has already vetted both, has social exposure if it goes badly, and can supply context that no profile contains. Removing that intermediary removed a filter that worked, and nothing on the platforms has replaced it.
The change is not confined to individual inboxes. Modelling by Josue Ortega and Philipp Hergovich, covered by MIT Technology Review, suggested that online introductions are changing the nature of society by creating links between people who share no social connections at all, with predicted effects on interracial marriage rates and on how long marriages last.
A woman’s difficulty in finding something serious is therefore happening inside a system that reorganised itself within a single generation, and nobody designed it for that outcome. The people who built it were solving for introductions, which it does extremely well, and not for what happens in month 4.
The Cost of Another Year
The cost is measured in years. A woman can spend a decade meeting a great many people under conditions engineered for continued use rather than for exit. Platforms are paid for retention; a user who leaves because she met somebody is a lost subscription, and no product team is rewarded for producing that outcome quickly.
Anybody who has been searching for more than 2 years without a single relationship reaching 6 months should treat that as information about the format rather than about herself. The places where a stranger arrives with a reference attached offer worse odds per encounter and considerably better odds per year. Nothing about another year of the same changes that arithmetic, and the years are not refundable.
