Why South Asian dating apps still feel like 2008.

This is the essay I have wanted to write for two years.

It is a critique of every major South Asian dating app on the market, written by someone who has used most of them, talked to dozens of people building them, and built the alternative to them. It will read as pointed. I am writing it that pointedly because the diaspora deserves a better critique of what is being shipped to it than the polite trade-press coverage these apps usually receive.

Three categories, three failures

The South Asian dating app market sorts into three categories, each of which is failing in a different way.

Category one: the matrimonial portals. Shaadi.com, BharatMatrimony, Jeevansathi. Their core product was designed for and shipped to an Indian parent generation in the late 1990s. They have skeumorphic I'd-like-to-print-this-out form layouts, family-broker chat models, photograph requirements that read as 2003-passport-photo, and a depth of cultural authority their successors lack. They are also, in user-facing UX terms, grimly antiquated. The current generation of marriage-age diaspora children will not use them, and the parents who would use them have nobody to set up. Their growth is held up by historical brand inertia in regional Indian markets, not by product strength. They are a product that aged into a museum.

Category two: the Western-style apps with a South Asian skin. Dil Mil, the diaspora-skinned Bumbles and Hinges, the various Tinder-with-Indian-options pivots. Their core product is swipe-and-meet. Their cultural specificity goes about as deep as the photo gallery and a "vegetarian only" filter. Their entire matching model is built on the assumption that the user is single-individual-with-no-family-in-the-decision, which is a model that is structurally incompatible with how the diaspora actually marries. They look like 2017 design with a sari emoji. They do not solve the problem.

Category three: the recent attempts at family-aware product design. Schmooze, a handful of South Asian seed-stage attempts in 2023–2024 that briefly trended on Y Combinator alumni Twitter, the few "matrimonial concierge" apps that have raised money in the last 18 months. Their problem is more sympathetic — they understand the cultural failure of the swipe model — but they have not yet figured out the actual product shape. Several have shipped buggy v1s, one has pivoted to consulting services, two have shut down. The intent was right; the execution has not yet caught up.

The common failure mode

What unites all three categories: they all start from the wrong question.

The matrimonial portals start from "How do we let Indian parents find marriages for their children online?" — which is a 1996 question, and the answer they built was for the 1996 internet. The product structure cannot evolve quickly because the user expectation that built it is now five product generations behind.

The Western-with-skin apps start from "How do we apply the swipe-and-meet model to Indian users?" — which is the wrong frame, because the swipe-and-meet model was never the right primitive for marriage-serious matchmaking in any culture. The Indian application of it inherits all of the original problems and adds the cultural mismatch on top.

The recent family-aware attempts start from "How do we make Indian parents feel involved?" — which is closer, but still wrong. Parents do not want to feel involved. Parents are involved, structurally, in the way Indian marriages have been made for centuries. The right question is not "how do we simulate the involvement"; the right question is "how do we give them their own account, with its own role, with its own boundaries."

What the right question actually is

I think the right starting question for a 2026-native South Asian matrimonial app is this: "What is the product surface for a family's matrimonial role, expressed natively in software?"

The question is structural. It does not ask about features. It asks about the shape of the product — what kind of account types exist, what kind of permissions they carry, what kind of artefacts they produce, what kind of conversations they enable.

The answers are not all obvious. They are answerable.

Account architecture. The Indian matrimonial tradition has more than one role. There is the Member who is the candidate. There is the family member who helps. There may be a third intermediary — the matchmaker-auntie, the wedding planner, the family friend who knows everyone. A 2026-native app needs account types that map to these roles. Not "main account + companion view." Distinct accounts with distinct authority.

Artefact preservation. The Indian matrimonial tradition has its own documents — the biodata, the kundali, the gotra lineage chart. A 2026-native app reads these directly, not as approximations but as the actual cultural artefacts they are. (See our guide to the biodata for what this looks like in practice.)

Compatibility framework. The Indian matrimonial tradition uses Ashtakoota Milan. A 2026-native app computes it as math between two specific charts, with no horoscope content, no daily readings, no astrology-as-engagement-hook. The framework is preserved in its original relational role. (See our Ashtakoota explainer.)

Cadence. The Indian matrimonial tradition is not a swipe stack. The pace is closer to one or two introductions per week, deliberated over by the family, considered with the candidate's career and life-stage in view. A 2026-native app should match that cadence — one curated introduction per day, capped server-side, no infinite scroll.

Cultural specificity in the headline. The Indian matrimonial tradition is a specific cultural practice with specific historical roots. A 2026-native app should declare itself as that — a matrimonial app for the Indian community, in the Vedic tradition, with the families participating. It should not pretend to be a generic platform that happens to be popular among South Asian users. The cultural specificity is not a bug to be optimised away; it is the load-bearing feature.

Why no current category leader can ship this

The reason none of the three categories can adopt the right structure: it would require reorienting their core, and their core is what they have raised money against. The matrimonial portals would need to ship a 2026 product, which contradicts the legacy expectations of their parent-generation user base. The Western-skinned apps would need to abandon swipe-and-meet, which is the only differentiator they have against Hinge and Bumble proper. The family-aware recent attempts would need to slow down to ship the right architecture, which contradicts the venture-pace they are operating at.

The structural opportunity, in other words, exists because the people in the best position to take it cannot afford to.

That gap is what Bhava is built into.

The case I am making

I am not making the case that Bhava has it all figured out. We have shipped v1.0; we have not shipped v3.0; the Concierge tier in the previous essay is a deliberate acknowledgment that the algorithmic side of the matching engine is not yet where it needs to be.

I am making the case that the direction is correct. Dual-account architecture. Cultural artefact preservation. Ashtakoota as math. One introduction per day. Cultural specificity declared in the headline. None of these are revolutionary; they are obvious in retrospect; they are what a Vedic matrimonial practice expressed in 2026 software should look like.

The current category leaders are not going to ship this. They cannot.

So the rest of us have to.

— Himanshu Batra