Showing Women Fewer Profiles Got Them Better Matches

Based on the research ofSabari Rajan Karmegam, Anandasivam Gopal, and Jui Ramaprasad, "Gender Gating? Addressing the Impact of Congestion on the User Experience for Women in Online Matrimonial Matching Platforms," Information Systems Research, 2026

Every matching platform designer inherits the same assumption: visibility is the product. Show a listing, a profile, a resume to as many relevant counterparties as possible, and the market will sort out the rest. A new field study of a leading Indian matrimonial platform, published in *Information Systems Research*, tests that assumption directly and finds it backward, at least on the side of the market that is already drowning in attention. Sabari Rajan Karmegam, Anandasivam Gopal, and Jui Ramaprasad worked with a platform to restrict which men could see a woman's profile at all, defaulting visibility to men who met commonly accepted norms on her age, education, and income. Women could switch the restriction off. Men could not. The result was not a quieter platform with the same outcomes. Women who were gated received fewer unwanted requests, converted far more of the requests they did receive into actual matches, and started reaching out to men themselves far more often than before, all without men's own outcomes getting worse.

The Market Congestion Problem Hiding in Plain Sight

Matrimonial and dating platforms built for Indian audiences share a structural feature that most Western dating apps also have, just more severely: a lopsided gender ratio. Karmegam has described the typical men-to-women ratio on these platforms as running between 60:40 and 90:10, and in the study's own data, women received on the order of 40 times as many expressions of interest as men did in the same period. That is not a market with too little activity. It is a market with too much of the wrong kind, concentrated on one side.

The behavioral logic driving the congestion is straightforward and, once stated, obviously self-defeating for everyone involved. When a man's odds of any given woman responding are low, the rational move is to contact many women rather than screen carefully before reaching out. Multiply that logic across a population skewed five or ten to one, and every woman's inbox fills with requests from men who were never a realistic match, sent not because he thought this profile was special but because sending costs him almost nothing and gives him one more roll of the dice. Karmegam has said the resulting deluge left women "extremely overwhelmed" in their first days on the platform, to the point that it obscured what the platform was even for. The paper connects that overwhelm to a familiar downstream outcome: reduced engagement and exit, women leaving a market that was ostensibly built to serve them because using it stopped feeling worth the effort.

This is not a phenomenon unique to Indian matrimonial sites; consumer dating platforms elsewhere show the same asymmetry in different clothing. Independent analysis of Tinder usage data finds women's average match rate near 44 percent against roughly 5 percent for men, alongside women receiving more than double the messages men do, a gap tied to indiscriminate swiping and messaging by the more numerous side. Pew Research Center's survey work backs this up qualitatively: six in ten women under 35 who have used a dating site or app say someone kept contacting them after they said they weren't interested, versus 27 percent of men the same age, and 57 percent of women 18 to 34 say they've received an explicit image or message they never asked for, roughly double the 28 percent of men who say the same. A 2023 scoping review in *Trauma, Violence, & Abuse*, pooling twelve studies, put the prevalence of unwanted sexual contact on dating apps between 57 and 88.8 percent, concentrated among women and sexual-minority users. Congestion like this is close to the default failure mode of any matching market where one side vastly outnumbers the other and messaging is cheap.

Screening at the Source Instead of the Inbox

Most platforms respond to this problem, if they respond at all, by giving the flooded side better filtering tools after the fact: block buttons, message requests, algorithmic ranking that tries to guess quality from behavior. Gender gating attacks the problem earlier in the pipeline. Rather than asking women to sort through everyone who reaches them, the platform decides in advance who gets to reach them at all, based on criteria the market itself already treats as relevant.

The mechanics are specific to the matrimonial context but generalizable in structure. The platform restricted visibility of a woman's profile to men whose age, education, and income fell within ranges consistent with prevailing social and institutional norms in Indian matrimonial matching, for instance, screening out a man in his forties from seeing profiles of women more than a decade younger or a few years older than himself. Women retained the ability to override their own default and open visibility further if they chose. Men had no equivalent override; if a woman's gate excluded him, he simply never saw the profile, and so never had the option to send her one of those low-cost, low-conviction requests in the first place.

The identification strategy is what makes the causal claim credible rather than merely correlational: the intervention rolled out in one subdomain of the platform while a comparable subdomain, matched on size and socioeconomic composition but differing in language and regional culture, served as the control, letting the paper isolate the effect of gating from the general trend of the platform or the season.

The intuitive fix for a flooded inbox is better filtering after messages arrive. Gender gating skips that step entirely, it decides who is allowed to send the message before congestion has a chance to happen.

What Restricting Visibility Actually Bought

The headline results run against the instinct that a platform should always maximize how many people can see a given listing. Treated women received measurably fewer incoming contact requests than the control group, congestion at the inbox fell rather than merely getting easier to manage. At the same time, matching efficacy, the rate at which incoming requests actually converted into a mutual match rather than being ignored or declined, rose by 72 percent. Fewer requests arrived, and a much larger share of the ones that did arrive were worth something.

The agency effect is arguably the more striking finding for anyone who assumed that restricting visibility would leave women more passive. Treated women initiated 75 percent of contacts themselves, a reversal of the typical pattern on a male-skewed platform where men do almost all of the outreach. With less time spent wading through requests that were never going to lead anywhere, women had the attention and confidence left over to reach out on their own terms, to men whose profiles the gate had already filtered for plausible fit, rather than a stranger cold-opening based on nothing.

The part of the design that made this viable inside the platform is what didn't happen: men's own outcomes did not worsen. Because the men who lost visibility into a given profile were, by the platform's own criteria, unlikely matches to begin with, removing that visibility mostly removed noise rather than opportunity. The platform judged the experiment successful enough to extend gating across its entire user base rather than keeping it confined to the treatment subdomain, the strongest signal available that this was not a one-off artifact of one region's data.

Where the Same Logic Travels

The George Mason University summary of Karmegam's research notes that the platform's move joins existing services such as JDate and SKY People, which already use culturally specific criteria as a sorting mechanism for their own communities. That is a reminder that norm-based screening on matching platforms is not a novel idea in principle, what this paper adds is a controlled measurement of what it actually buys the constrained side, and confirmation that the unconstrained side does not pay for it.

The broader applicability is easiest to see wherever a platform combines a badly skewed population ratio with near-zero marginal cost of reaching out. Karmegam has floated ride-hailing driver pools and academic hiring pipelines as analogous settings, and it isn't hard to extend the list to cold outreach on recruiting platforms or any marketplace where one side can blast messages faster than the other side can evaluate them. The matrimonial context is unusually well suited to gating because "accepted social norms" on age, education, and income already function as a widely shared, low-controversy screen inside that market, a condition that won't transfer cleanly everywhere, and operators borrowing the idea should expect to spend real effort finding criteria their own users will accept as legitimate rather than arbitrary.

The Indian matrimonial setting also matters for reading the results honestly rather than as a universal law. Family involvement in match selection remains the norm rather than the exception here, national survey data cited by the BBC put the arranged-marriage share of Indian marriages at 93 percent as of a 2018 household survey of more than 160,000 respondents, and matrimonial platforms in this category are frequently used jointly by prospective partners and their parents, who apply their own filters on top of the platform's. Gender gating layers a platform-level norm filter onto a market that already runs on norm-based filtering by families. Whether the same intervention would produce a comparable efficacy gain in a market with far less appetite for criteria-based screening, such as a Western casual-dating app, is an open question the paper does not claim to answer.

What the study does establish, cleanly, is that the assumption underlying most platform design, that expanding who can see a listing is close to a free win, breaks down once one side of the market is already saturated with low-quality attention. In that condition, the platform's real lever isn't reach. It's precision about who gets access in the first place, and a willingness to take that access away from the men whose messages were never going to land regardless of how many of them got sent.