The data is in. The oldest system still wins. Here is what the numbers actually say.
The Chassidic world has no shidduch crisis.
That is not an opinion. It is a demographic fact that anyone who has spent time in these communities already knows, and everything else in this article makes sense only in relation to it.
In most Chassidic communities, a boy does not meet a girl without exhaustive research first. His family investigates her family. Her family investigates his. Neighbors are called. Teachers are consulted. People who davened next to the father for twenty years are asked what kind of person he is. The Rav is involved. The research can take weeks, even months. Only after all of that does a meeting happen. The meeting itself is brief — usually one meeting to reach a decision. Many marry the first person they meet. A Chassidic boy would not attend a singles mixer any more than he would eat treif. A mixed event is not one option he decided against. It is simply outside the framework of how his world works.
Both boys and girls enter shidduchim at eighteen. There is no age gap. In Skver, each year a list is made of all the boys and girls of marriageable age, and the shadchanim go down the list. Within a year or two, virtually everyone is married. Not because they have no expectations, but because the research was done before they ever sat across from each other. By the time the boy walks into that room, the question is not “who are you?” The question is “can I build a life with you?”
Alexander Rapaport, a Hasidic father of six and executive director of Masbia, was once asked about the shidduch crisis. He seemed perplexed. He said he had heard of it but was not sure he understood what it was about. He was not being obtuse. He was describing his reality. In his world, it would be strange for a man to marry a woman even two years younger than him. Everyone starts at the same time. The shadchanim do their work. The system moves.
If anything, the Chassidic world has the reverse problem. Because both genders start at eighteen and roughly a hundred and five boys are born for every hundred girls, there are slightly more boys than girls in the pool. The NASI Project documented this — a Chassidic boy’s crisis, not a girl’s crisis. But it is not really a crisis at all, because the system absorbs the imbalance quickly. The structure works. Shidduchim, practiced in its purest form — without apps, without mixers, without self-selection of any kind — produces marriages.
The Gradient
Move one step along the spectrum, from Chassidish to Yeshivish, and a problem appears. Girls start shidduchim at nineteen. Boys start at twenty-two or twenty-three, after years in Beis Medrash and often a year or two in Eretz Yisrael. In a community that grows at roughly four percent a year, every cohort of a thousand boys at twenty-three faces a cohort of about eleven hundred and sixty girls at nineteen. The math does not care about anyone’s feelings. This is the age-gap theory, and while the DAAS study found the actual marital age gap is closer to two years than four, the mismatch in entry timing creates real pressure.
Even so, the yeshivish system still works remarkably well. Ninety-two percent of both men and women are married by thirty. By forty, ninety-eight percent. These numbers come from the largest study of its kind — nearly nine thousand individuals. They are extraordinary by any measure. In the general U.S. population, only twenty percent of adults aged eighteen to twenty-nine are married today. The yeshivish world, for all the hand-wringing, is still producing marriages at a rate the rest of America cannot remember.
Move another step, to Modern Orthodox. Eighty-one percent married by thirty. Just under ninety percent by forty. Still far better than the national average. But the ten-point gap between yeshivish and Modern Orthodox is not a coincidence. It tracks almost perfectly with the degree to which self-selection — dating apps, mixers, singles events — has supplemented or displaced the shadchan.
Move further still, to the broader American Jewish community where dating apps are dominant and shadchanim are largely absent, and the pattern mirrors the national collapse: marriage rates down sixty percent since 1970. Median age at first marriage the highest ever recorded — thirty point eight for men, twenty-eight point four for women.
The more purely a community relies on shidduchim, the higher its marriage rate. The more it shifts toward self-selection, the lower the rate goes. This is not a commentary on anyone’s frumkeit. It is a structural observation about which system produces the outcome that everyone says they want.
The Alternatives and Their Numbers
There is no shortage of ways for Jewish singles to meet today, and the numbers tell a story.
JWed reports over thirty-seven thousand marriages since 2001. JDate claims responsibility for fifty-two percent of Jewish marriages that began online and has over seven hundred and fifty thousand members. SawYouAtSinai, which combines profiles with three hundred-plus matchmakers, reports over four thousand engagements from around thirty thousand members. JSwipe, Lox Club, Forj, Hinge with a Jewish filter — there are more platforms than anyone can keep track of.
Corona Crush, which started during lockdown as a Facebook group, has grown into the largest Jewish speed-dating operation in the world. Over a hundred and eighty events, averaging five hundred participants each. A 2025 event connected nearly seven thousand singles in a single session, generating over fifteen thousand dates. They report a hundred-plus marriages. Met@Chabad matched four thousand singles across two hundred and fifty Chabad houses. YUConnects has facilitated over three hundred and fifty engagements. Jewish Dating NYC claims forty-five engaged couples. Add Shabbatons, singles weekends, wine tastings, game nights, mixers in Manhattan apartments, and the informal networks of rabbis and rebbetzins who set people up over Shabbos tables.
In Israel, the landscape is just as crowded. Shagririm Ba’Lev — Ambassadors of the Heart — started as a student project at the Jerusalem College of Technology and has become one of the most talked-about initiatives in the dati leumi world. Three hundred and forty-five couples married since 2019, with a new couple getting engaged roughly every two days. Their model is built on ambassadors — friends who register their single friends and advocate on their behalf, searching a database of twelve thousand singles. Twenty-five percent of the candidates who joined have gotten married through the initiative, and their date-to-marriage average is three and a half fewer dates than other platforms. GamZuli matches observant singles based on a detailed questionnaire, with the first meeting a fifteen-minute video chat. Shiduch.org connects singles with shadchanim across Israel, Europe, and the U.S. And of course, the secular Israeli apps — Tinder, Bumble, OkCupid — have enormous market share among non-Orthodox Israelis.
By any measure of activity, the ecosystem is thriving. In Israel and America both, there has never been more infrastructure dedicated to getting Jewish singles to meet each other.
But activity and outcome are not the same thing.
Corona Crush’s hundred-plus marriages from a hundred and eighty events averaging five hundred people means roughly one marriage per event. One outcome for every five hundred attendees. Their fifty-percent mutual-match rate sounds encouraging until you track how many of those matches become second dates, and how many second dates become engagements. The funnel narrows dramatically. Pew Research found that only twelve percent of Americans who have used a dating app ended up married or in a committed relationship. Twelve percent. In an industry generating eleven billion dollars a year.
Compare this to Skver, where they go down a list and within a year or two, everyone is married.
What Each Model Actually Is
Every dating method has a theory of how marriages happen, whether or not it says so.
A dating app — JDate, JSwipe, Lox Club, Hinge — operates on self-selection. You see a profile. You make a judgment based on photos, a bio, maybe a few prompted answers. You swipe or you don’t. If both swipe, a conversation opens. From there, you are entirely on your own. No one is guiding the process. No one knows both sides. No one is invested in the outcome.
Speed dating and events — Corona Crush, Jewish Dating NYC, Shabbatons, mixers — operate on curated self-selection. You are placed in a room with people who meet some basic criteria, and from there it is still self-selection. You have five to eight minutes to judge. The decision — yes or no, interested or not — is still yours, made under pressure, in a compressed timeframe, with very limited information.
A hybrid platform — SawYouAtSinai, YUConnects — adds human judgment to a digital system. A matchmaker reviews your profile, considers the pool, sends suggestions. This is the closest thing to traditional shidduchim operating at scale, and its outcomes reflect that. Four thousand engagements from thirty thousand members is a conversion rate no pure self-selection platform can touch.
And then there is shidduchim. A shadchan who knows you, knows the other person, picks up the phone, and makes the case. Who remembers that the girl from Yerushalayim mentioned, almost in passing, that she grew up near the ocean and misses it. Who notices that the boy’s mother keeps steering the conversation back to middos and away from parnassah, and understands what that means. Who can sit across from someone for twenty minutes and come away knowing something that no profile and no algorithm and no seven-minute speed date could ever capture: what this person is actually like when they are not trying to present themselves.
For a Chassidic boy or a very yeshivish girl, this is not one option among several. It is the only option. Sitting in a room evaluating strangers of the opposite gender over plastic cups of wine is, in their world, simply not done. And their marriage rates are the highest of any group.
Follow the Money
A dating app needs you to keep dating. This is not cynicism. It is arithmetic. Tinder processes two billion swipes a day and earns one point six billion dollars a year. Bumble earns nine hundred million. NPR’s Planet Money called this the dating app paradox: the app’s success is measured by how many users it retains, but the user’s success is measured by leaving. Every successful match costs the platform two paying customers.
JDate charges around sixty dollars a month. JSwipe has premium tiers. JWed has subscription options. The economics are the same: a married person cancels. An engaged person stops paying. The business model profits from the search continuing.
An event organizer — whether charging tickets or running on donations — profits from running events, not from producing marriages. More events, more revenue. The longer people remain single and keep attending, the larger the audience. Nobody is planning this intentionally. But the structure is the structure.
And here is what people miss about how this affects the suggestion itself. A platform that profits from activity has every reason to send you another match, another notification, another name — whether or not it is any good. Volume is the point. A bad suggestion is not a failure for them. It is another interaction, another reason for you to log in tomorrow.
A shadchan gets paid when there is a shidduch. Not before. Not per date. Not per month. Not per event attended. At the engagement. A shadchan who does not produce results earns nothing. Every month that passes without a match is a month the shadchan works for free. The entire incentive structure points in exactly one direction: get this person married.
This is not a minor distinction. It is the foundational difference between a system designed to sustain a process and a system designed to produce an outcome.
The Illusion of Control
So if the incentives are wrong and the outcomes are worse, why do people keep choosing the alternatives? The answer is not complicated. They feel like you are doing something.
Swiping is active. Browsing profiles is active. Choosing which event to attend, which app to download, which filters to set — all of this feels like agency. You are in the driver’s seat. You are making decisions. You are taking your life into your own hands. And waiting for a shadchan to call feels like the opposite. It feels passive. It feels like sitting by the phone. It feels like surrender.
But this is precisely backwards. When someone else curates, you evaluate. When you curate for yourself, you search — and searching becomes the activity. The app does not ask you to decide about one person. It asks you to decide about hundreds, one face at a time, in a never-ending scroll. The event does not ask you to be present with one human being. It asks you to scan a room. The psychology of this is well established: the more options you process, the less capable you become of choosing any of them. Barry Schwartz called it the paradox of choice. It is not a theory. It is what happens to people in laboratory after laboratory, and it is what happens to singles on app after app.
The feeling of control is the product being sold. Not marriage. Not even dates. The feeling that you are doing everything you can. And the more you do, the more you need to keep doing, because stopping feels like giving up. A person on three apps who attends two events a month and has a profile on a matching site is not three times more likely to get married than a person doing one of those things. They may be less likely, because their attention is fractured, their expectations are confused by overexposure, and their emotional energy is spent on the process of searching rather than the work of connecting.
Shidduchim does not feel like control because it is not control. It is delegation. You delegate the search to someone whose livelihood depends on getting it right. You delegate the initial filtering to people who know both families, who have context you cannot get from a photograph and a paragraph. And when a name comes to you, you are not choosing from a catalog of five hundred. You are considering one person, seriously, with your full attention. That is not passivity. That is the opposite of passivity. It is the most active form of discernment there is — the kind that leads to a decision rather than another round of browsing.
The model that feels like agency produces paralysis. The model that feels like waiting produces marriages.
It Is Always a Shidduch
There is a common misconception that shidduchim only happens through professional shadchanim. That if your neighbor’s wife mentioned a name over Shabbos lunch, or your friend from seminary thought of someone, or your uncle heard about a boy from his chavrusa — that somehow this is not really a shidduch. That it is informal. That it does not count.
It counts. By halacha, it is a shidduch. If someone — anyone — suggests a match and that suggestion leads to a marriage, shadchanus is owed. This is not a technicality. It is a recognition that the act of bringing two people together is work that deserves compensation, regardless of whether the person who did it carries a business card. The neighbor at the Shabbos table exercised judgment. She thought of two people, considered whether they might be right for each other, and said something. That is what a shadchan does. The halacha does not distinguish between the professional and the neighbor. It distinguishes between someone who facilitated a match and someone who did not.
This matters because it means the system is larger and more intact than people realize. When people say “shidduchim is dying,” they usually mean that fewer professional shadchanim are active, which is true. But shidduchim — the act of one person seeing two other people and recognizing a possibility — happens every day. At kiddushim. Over the phone. In WhatsApp messages between mothers. The infrastructure is strained, but the instinct is alive. Every time someone says “I know someone who might be perfect for your son,” that is the system working. And the halacha says it is owed the same respect and the same compensation as a match made by the most established shadchan in Flatbush.
Does the Shadchan Model Still Work?
The incentive structure is sound. The family involvement is right. The constrained choice — one suggestion at a time, considered seriously — is exactly what the research says produces better decisions. Schwartz demonstrated that more options produce more anxiety, less satisfaction, and a decreased ability to commit. That describes dating apps precisely. And it describes what a good shadchan prevents.
But “sound in theory” and “working in practice” are not the same thing.
The OU surveyed twenty-three hundred single Orthodox men and women and interviewed nearly a hundred shadchanim and community leaders. The findings were unambiguous. Singles highly dislike using shadchanim. They find the advice demeaning, the suggestions off-base, the process dehumanizing. Shadchanim, for their part, feel unfairly compensated, overwhelmed, and unappreciated. The OU identified two distinct crises: a Crisis of Process — the mechanics are broken — and a Crisis of Experience — the process itself is doing harm.
When we started Yismach, we gave access to the established, recognized, professional shadchanim — the people everyone knew, the names that came up whenever someone asked “who should I talk to?” At its peak the English site had over two hundred shadchanim on the roster. A few years later, we looked at the activity logs. Many of those well-known, respected shadchanim — some of them famous in their communities — had either stopped being shadchanim entirely or stopped using the platform. The roster came down to about a hundred. Today, limited to those who are active and consistent, it is about forty.
This is not a condemnation of those who left. Being a shadchan is thankless work, and more often than not it is humiliating. You invest hours in a suggestion and get told no before the other side has even heard the name. You bring two people to a fourth date and one of them disappears without a word. A family you helped forgets you exist the moment the l’chaim is over. People walk away from that. It is understandable.
The system with the best structural design is also the system hemorrhaging the people who make it run. The architecture is right. The execution is breaking down.
What We Are Doing About It
The OU diagnosed two problems: a broken process and a damaging experience. We built Yismach to address both.
On the experience side: the most common complaint is that shadchanim are demeaning or dismissive. This is not a perception problem. It happens. And it has to stop. We maintain active constraints for every shadchan on the platform. A shadchan who is reported as insulting or offensive is removed. Not warned. Removed. A shadchan who charges for initial consultations — for intake meetings, for getting-to-know-you conversations, for work that the shadchanus itself is supposed to cover — is removed. A shadchan who runs competing self-selection channels — mixers, gatherings that operate on “look around the room and figure it out yourselves” — is removed. A shadchan who is inactive for an extended period is removed. We went from two hundred to forty not by accident, but by decision about what this platform is and who belongs on it.
Privacy is not a feature. It is the foundation. Your profile on Yismach is not browsable by other singles. There is no swiping. There is no catalog. Your information is seen by screened, active, accountable shadchanim and by no one else. When a suggestion is made, it goes through the shadchan. The families are involved at the appropriate time. Nothing is public. Nothing is exposed. For singles who have spent years on platforms where their photos are visible to thousands of strangers, where a rejected match can screenshot their profile and send it to a group chat, where they feel like merchandise in a storefront window — this is not a small thing. It is the difference between being known and being displayed.
The Process
The OU found that suggestions are often off-base. This is worth thinking about carefully, because it touches on what a good suggestion actually is — and what happens when you send one that is not.
A shadchan who knows three hundred people is exceptional. But three hundred is a fraction of the eligible pool. The shadchan in Bnei Brak does not know the family in Teaneck. The shadchan in Flatbush has no visibility into profiles entered by a shadchan in Yerushalayim. With thousands of profiles in the system, there are millions of possible pairs, and no human being can evaluate them all.
This is what our proprietary AI addresses. Not the judgment. Not the human part. The visibility. The AI analyzes profiles across the entire database and identifies patterns of compatibility that no single shadchan could hold in their head. It crosses geographical boundaries, crosses the invisible walls between shadchanim who work different segments, and surfaces possibilities that would otherwise go unseen.
But the AI does not just surface names. It produces a match comparison — a detailed breakdown of where two people align and where they differ, across hashkafa, middos, family background, life goals, what they said they want and what their profile actually reflects. The shadchan sees this breakdown before picking up the phone. And when the suggestion reaches you or your family, you see it too. Not a name and a resume. A reason. This is why we think this person could be right for you. Here is where the alignment is strong. Here is where there is a difference worth discussing. Here is what the shadchan saw that made them pick up the phone.
This changes what a suggestion means. And it changes when a suggestion gets made.
On an app, you get a notification: someone liked your profile. At a mixer, you get seven minutes. On a subscription platform, you get another name in your inbox because the algorithm needs you to stay engaged. The economic pressure in all of these models is to keep giving you something — anything — so you stay in the pipeline.
On Yismach, if there is no good suggestion, we do not make one.
We would rather you hear nothing for a month than hear a suggestion that wastes your time and erodes your trust in the process. Because that is what bad suggestions actually cost. Not just an awkward date. Trust. A single who gets three off-base suggestions in a row stops believing the shadchan knows what they are doing. They stop returning calls. They start drifting toward the apps and the events, not because those are better, but because at least there nobody is pretending to know them.
The match comparison makes this possible. Without it, the shadchan is working from memory and instinct — and even a great shadchan’s memory has limits. With the breakdown in front of them, they can see specifically where the alignment is and where it is not. They can make an informed decision about whether this suggestion is worth making. And when they do make it, you can see the reasoning. Not “trust me, I know someone.” But “here is what I see, and here is why I think it is worth your time.”
The AI gives the shadchan the reach. The match comparison gives the clarity. The shadchan provides the judgment.
What Actually Predicts a Marriage That Lasts
The goal is not just to get you to the chuppah. It is to get you to a marriage that endures.
The same four things predict whether a marriage is happy and whether it lasts: a deep personal commitment to the marriage as an institution, a feeling that your spouse is protective of the relationship, regular religious practice together, and consistent investment in time as a couple. Not chemistry. Not attraction. Not how many dates you went on before you decided. Commitment, protectiveness, shared worship, and showing up. The same four, for husbands and wives alike.
Decades of observation at the University of Washington found that couples in stable marriages respond to each other — turn toward each other when the other reaches out — eighty-six percent of the time. Couples who eventually divorce? Thirty-three percent. Stable marriages maintain at least five positive interactions for every negative one. They bring up disagreements gently instead of with contempt.
A systematic review across multiple countries found the same things surfacing everywhere: commitment, loyalty, trust, communication, empathy, shared values, and support from the extended family.
Not one of these is something you can evaluate from a dating profile. Not one shows up in a five-minute speed date. Not one is captured by the swipe. Commitment, protectiveness, the capacity to turn toward your spouse when it is easier to turn away — these reveal themselves slowly, in context, under pressure. They are exactly the kinds of things a shadchan who has sat with both families, who has watched how they respond to disappointment and uncertainty, who knows what the mother values and what the father worries about, is positioned to assess.
And notice what the shidduch system embeds by design. Family involvement — one of the strongest predictors of marital quality. Shared religious practice — top four predictor of both happiness and stability. Realistic expectations — because a shadchan does not promise you your soulmate, a shadchan introduces you to someone you could build a life with. And constrained, intentional choice — one suggestion at a time, considered seriously, given real attention — which is precisely the decision structure that produces better outcomes.
The Chassidic community did not read any of these studies. They did not need to. They have been doing this for generations because the mesorah said to, and it turns out the mesorah was right.
The Chassidic world, which uses shidduchim exclusively and would not consider any alternative, has no crisis. The yeshivish world still marries at ninety-two percent by thirty. The Modern Orthodox world, where self-selection has made significant inroads, marries at eighty-one percent. The broader American Jewish community, where dating apps dominate, increasingly mirrors the national trend.
The alternatives have been tested. The numbers are in. Seven hundred and fifty thousand JDate members. Thirty-seven thousand JWed marriages in twenty-four years. Fifteen thousand Corona Crush speed dates in a single evening. Eleven billion dollars in annual dating industry revenue. A twelve percent success rate for app users. And the lowest marriage rates in American history.
Meanwhile, in Skver, they go down the list.
The problem is not that shidduchim does not work. In its purest form it works better than anything else that has been tried. The problem is that the infrastructure outside the Chassidic world is eroding. Shadchanim are leaving.
The ones who remain are overwhelmed and undercompensated. The culture is drifting toward self-selection because it feels modern and empowering, even as the data says it produces more activity and fewer marriages.
But the system is more intact than people think. Every time a neighbor mentions a name, every time a friend from seminary thinks of someone, every time an uncle hears about a boy from his chavrusa — that is shidduchim. It is not informal. It is not casual. By halacha, it is a shidduch, and shadchanus is owed. The professional infrastructure may be strained, but the instinct to see two people and recognize a possibility — the instinct that has sustained Jewish marriages for thousands of years — is alive in every community, at every kiddush, in every WhatsApp group of mothers who cannot stop themselves from thinking about who might be right for someone else’s child.
We built Yismach because someone has to strengthen that infrastructure. Not replace the shadchan with technology. Not add another self-selection platform to the pile. Strengthen the shadchan. Give the shadchan AI tools that extend their reach across the full database. Give them — and you — a clear breakdown of why a suggestion makes sense, so that every suggestion that goes out is one worth making. And when there is nothing worth suggesting, have the discipline to say nothing. Enforce constraints that protect singles from a demeaning experience. Ensure privacy that honors the dignity of the process. And preserve the one incentive structure that actually works: the shadchan gets paid when there is a shidduch. Not for activity. Not for volume. For the outcome.
The oldest matchmaking system in the Western world is not a relic. The community that practices it most faithfully is the community with the highest marriage rate and no crisis to speak of. The research on what makes marriages last reads like a description of what shidduchim already does. The question is not whether the system works. The question is whether we are willing to invest in making it work beyond the communities where it never stopped.
• • •
Notes and References
1. Chabad.org, “The Shidduch: How Jewish Dating Works.”
2. The Yeshiva World Forums, discussion of Skver’s shidduch system.
3. Jon Birger, “Dateonomics” (2015); Alexander Rapaport interview.
4. U.S. Census Bureau. Approximately 105 males born for every 100 females.
5. NASI Project (shidduchproject.com).
6. Comenetz research; reported in Birger, “Dateonomics” and Ami Magazine (2013).
7. Dr. Yosef Sokol et al., DAAS study, JSSR (2022). Average marital age gap: 2.2 years.
8. DAAS study (Sokol et al., JSSR 2022). Yeshiva Orthodox: 92% married by 30; ~98% by 40.
9. U.S. Census Bureau, Current Population Survey. First marriage rate fell ~60% since 1970.
10. Pew Research Center (2024). Adults 18–29 married: 20% vs. 59% in 1960.
11. U.S. Census Bureau, Historical Marital Status Tables (2024). Median first marriage: 30.8 men, 28.4 women.
12. JWed.com; Top Consumer Reviews (February 2026).
13. JDate via DatingNews.com (November 2025); Hadassah Magazine (November 2022).
14. SawYouAtSinai.com; Innerbody.com (January 2025).
15. CoronaCrush.co; The Jewish Link (May 2024).
16. YNet News (August 14, 2025).
17. Corona Crush Instagram (@coronacrushofficial).
18. Lubavitch.com (May 2021); MetatChabad.com (August 2021).
19. Jewish Action (OU, August 2022); Dr. Efrat Sobolofsky, YUConnects.
20. JewishDatingNYC.com.
21. CoronaCrush.co event results.
22. Pew Research Center (2023). 12% of dating app users ended up married/committed.
23. Global dating industry revenue: $11.4 billion (2024). Statista.
24. SawYouAtSinai: 4,070 engagements from ~30,000 members.
25. Match Group 10-K; Bumble Inc. financials.
26. NPR Planet Money, “The Dating App Paradox.”
27. DatingNews.com (November 2025). JDate: ~$60/month.
28. Barry Schwartz, “The Paradox of Choice” (2004).
29. OU research study, reported by 18Forty.
30. Institute for Family Studies & Wheatley Institute (2022). Four predictors of marital happiness and stability.
31. John Gottman, “The Seven Principles for Making Marriage Work” (1999). 86% turn-toward rate; 5:1 ratio.
32. BMC Public Health (2019). Systematic review of protective factors in long-term marriage.