100,000 Matchmaking Clients Reveal the Age-Gap Divide Apps Ignore

The cultural conversation about age-gap relationships has been running loud for years. The numbers haven't moved. A Tawkify survey of 98,798 paying matchmaking clients across all 50 US states shows the gendered asymmetry in age preferences holding firm in 2024: men date younger at nearly twice the rate they date older, women reverse that ratio with almost mathematical precision, and the gap between those preferences is wide enough to materially affect who sees whom on any platform using age filters as primary sorting logic.

These aren't casual swipers responding to a pop-up survey. They are people who paid a professional matchmaking service to find them a partner. The persistence of this pattern at that level of deliberate intent is what makes it worth examining, not as a moral question, but as a product architecture problem that the dating industry keeps setting aside.

The High Intent Take

The gendered asymmetry in age preferences is not a bug in your user base. It's a structural feature of the market you're building for. Nationally, men cap their older-dating ceiling at 8.76 years while women push their floor to 14.12 years older. Those ranges do not overlap cleanly, and when your algorithm sorts by age filter, you are amplifying the mismatch rather than bridging it.

The move for product teams is not to override user preferences. It is to understand that stated preferences and revealed preferences diverge when supply gets thin, and to design systems that surface that divergence honestly. Apps that quietly expand radius when local pools run dry already know this. Age deserves the same nuanced treatment. Build the data infrastructure to understand where your users actually match versus where they say they will, then build features around that delta.

The Numbers, State by State

Nationally, 95% of women in Tawkify's sample said they were open to dating older partners, with an average acceptable age gap of 14.12 years. Among men, only 65.7% said the same, with an average ceiling of 8.76 years. On the other side: 96.5% of men would date younger (average gap: 14.7 years), compared to 88.1% of women at a much narrower average of 7.14 years. That is not a marginal difference. It is a systematic divergence reproduced across every state in the country.

Alaska topped the charts for women's openness to older partners, 93.1% open, with an average acceptable gap of 20.57 years. Delaware recorded the highest average gap for men willing to date older at 11.26 years, while Wyoming had the highest male openness rate at 71.7%. For younger partners, Hawaii led for both genders: 96.8% of men were open to it at an average gap of 18.59 years, and 92% of women at an average of 9 years. Hawaii's median age of 41.5 years, compared to the national 39.2, may be part of that picture, or it may simply reflect the mathematics of a geographically isolated dating market where flexibility is a survival mechanism.

The states showing the most flexibility on age also tend to be the least populated. Alaska, Wyoming, North Dakota, these aren't markets with deep benches. Stated preferences bend under supply constraints, and operators already know it.

That last point matters for product design. Tinder will quietly show you profiles outside your stated radius if the local pool runs dry. The same logic applies to age. Users say they won't consider someone ten years older; when the alternative is no matches at all, the calculus shifts. The question is whether you design for the preference users state or the preference they reveal under constraint.

Why the Cultural Reckoning Hasn't Arrived

Large age gaps have faced more social scrutiny in recent years, particularly male-older pairings. The discourse around power imbalances, life-stage mismatches, and the dynamics of older men dating significantly younger women has moved from academic circles into mainstream dating culture. None of that has changed the underlying preference data in Tawkify's sample.

The gendered asymmetry maps onto decades of research on mate selection, status, and perceived fertility. It also maps onto the economic realities of gender and aging: men's earnings and social capital tend to peak later and decline more slowly than women's perceived social desirability as partners. Dating apps don't create those dynamics. They do amplify them. Age is one of the few datapoints that is both universally collected and universally filtered on, and most platforms display it prominently. That makes age-based sorting faster and more brutal than it would be in an offline context where other cues, warmth, humor, presence, might soften the initial calculus.

Research shows that around 8% of male-female couples in Western countries have an age gap of 10 years or more, yet the polling data on preferences consistently shows much higher stated openness to large gaps, particularly from women toward older men. The gap between stated openness and actual couple formation suggests that willingness is there but that platforms, social context, or self-selection filters it out before it converts to relationships.

What This Means for Product and Match Architecture

Hinge, Bumble (BMBL), and Match Group's (MTCH) portfolio apps all let users set age ranges, and the Tawkify data suggests those ranges are being set in systematically gendered ways with predictable consequences for who sees whom in each feed. Women set wide upper bounds and narrow lower bounds. Men set wide lower bounds and narrow upper bounds. The resulting pool asymmetry is structural, not random, and it compounds over time as users get frustrated with match quality without understanding why.

Most apps let users filter aggressively, then wonder why match rates are low. The data suggests users will compromise on age when they have to, but they won't do it voluntarily, and current filter design gives them no reason to try.

A handful of platforms, Thursday, Feeld, The League at various points, have experimented with restricting filters to force broader consideration. The trade-off is always autonomy versus efficiency, and autonomy usually wins in user research because nobody wants to be told they can't filter. But the Tawkify data creates a different framing: what if the filter itself is what's reducing your chances? If a woman's stated age-gap tolerance of 14.12 years toward older partners is real, and a man's openness ceiling of 8.76 years toward older partners is real, the overlap zone is identifiable. Build a feature that shows users where they're being excluded from mutual-interest pools, and let them opt into wider consideration with that information in hand.

Tawkify's survey is promotional research and ends, predictably, with reassurances about communication and shared values. The numbers deserve more credit than the framing. The dating industry has run hard at inclusivity, representation, and bias reduction for the better part of a decade. Age is the frontier where almost none of that work has happened. The preferences are durable, the asymmetry is wide, and the match-rate consequences are real. That is the conversation worth having.