Yidui's IPO: Is It a Dating App or a Live-Streaming Giant?

A Chinese dating platform that deploys nearly 200,000 professional hosts to sit in on video dates has tripled its profits and is heading for a Hong Kong IPO with financials that dwarf Western competitors on key metrics. Yidui's conversion rate of paying users reaches 16.4%—more than three times what Tinder achieves—but the revenue model looks less like matchmaking and more like live-streaming entertainment with romantic overtones. The company has now refiled for a Hong Kong IPO after withdrawing an earlier attempt in late 2023.

The mechanics are straightforward: two users interested in each other jump on a three-way video call with a professional host who facilitates conversation, breaks silences, and generally plays digital wingman. Users pay for this service overwhelmingly through virtual gifts—think live-streaming economics grafted onto dating—rather than subscriptions. Virtual gifts accounted for 98.7% of revenue in fiscal 2024, with subscription revenue barely registering at 1.3%, or roughly $7.4M.

What's striking isn't the model itself—live-streaming platforms across Asia have used paid hosts for years—but the scale and the conversion rate. Yidui's 16.4% paying user base compares favourably to Western swipe-based platforms, where conversion typically hovers between 3% and 5% for premium tiers. Tinder's conversion to paying subscribers sits at roughly 4%, according to Match Group's most recent disclosures, whilst Bumble operates in a similar range based on industry estimates.

Create a free account

Unlock unlimited access and get the weekly briefing delivered to your inbox.

Register free

No spam. No password. We'll send a one-time link to confirm your email.

This is either the most successful dating product innovation of the decade or proof that live-streaming entertainment in a dating wrapper monetises better than actual matchmaking.

The 16.4% paying rate tells you something works—but the 98.7% virtual gift revenue split tells you what's working might not be romance. If hosts are incentivised by prolonged interaction rather than successful matches—and the filing doesn't clarify commission structures—then Yidui has built a brilliantly effective entertainment business that happens to involve two people who might fancy each other.

Revenue model raises definitional questions

The virtual gift economy presents a challenge for anyone trying to benchmark Yidui against dating peers. Live-streaming platforms like Momo (owned by Hello Group) or Douyin operate on identical revenue mechanics: users purchase virtual items and send them to hosts or performers during live broadcasts. Hosts take a commission, the platform takes a cut, and engagement drives revenue—not outcomes like matches, dates, or relationships.

Yidui describes its hosts as professionals trained to 'help shy or socially awkward users break the ice and facilitate meaningful conversation', according to the IPO prospectus. That framing positions the service as dating enablement. The revenue data suggests users are paying for company, entertainment, or validation—outcomes more commonly associated with live-streaming or even companionship services than traditional dating apps.

The question isn't whether this model works. Clearly it does—Yidui's average revenue per paying user clocked in at $174 for fiscal 2024, up from $145 the previous year. The question is whether this is a dating company or a live entertainment platform with romantic subtext. That distinction matters for operators elsewhere considering whether human facilitation could address retention and monetisation challenges in swipe-fatigued markets.

Commission structures and the incentive problem

The IPO filing discloses that Yidui's 193,000 hosts work on commission but doesn't specify how that commission is calculated. Are hosts rewarded for successful matches? For time spent on calls? For the volume or value of virtual gifts sent during sessions? The structure determines whether hosts are aligned with user dating outcomes or with maximising session length and gift-sending behaviour.

If the latter, Yidui has effectively created a gig economy of professional third wheels whose income depends on keeping users engaged and spending, not on whether those users eventually meet, date, or delete the app. That's not necessarily problematic—plenty of businesses monetise attention rather than outcomes—but it does make the product categorically different from platforms that charge subscriptions tied to profile visibility, unlimited swipes, or advanced filters.

Yidui has effectively created a gig economy of professional third wheels whose income depends on keeping users engaged and spending, not on whether those users eventually meet, date, or delete the app.

The model also raises obvious trust and safety questions. Yidui mentions 'romance scams' as a disclosed risk but provides no incident data, mitigation protocols, or specific safeguards beyond general platform monitoring. A three-way video format theoretically reduces certain scam vectors—it's harder to catfish when a third party is present and video is mandatory—but introduces others. Hosts could coordinate with one party to extract gifts from the other, or users might develop parasocial attachment to hosts rather than matches.

Market context: Hello Group stumbles, AI rivals push different thesis

The IPO timing benefits from turbulence at Yidui's primary competitor. Hello Group, which operates Momo and Tantan, disclosed in January 2025 that Chinese tax authorities are investigating the company for potential underreporting of income between 2020 and 2024. Hello Group's user base has also declined—Momo's paying user count dropped to 8.8 million in Q3 2024, down from 9.6 million a year earlier.

Meanwhile, Soulgate—another Chinese dating platform that filed for a Hong Kong IPO in late 2024—has taken the opposite strategic bet. Soulgate emphasises AI-driven features, virtual avatars, and asynchronous interaction. Where Yidui bets on human facilitation, Soulgate is automating connection. Both companies are essentially running controlled experiments on what Chinese users want from dating platforms: high-touch human mediation or high-tech algorithmic matching.

The contrast matters beyond China. Western operators watching both models could draw different lessons. Yidui's success suggests swipe fatigue and low engagement might be solvable through labour-intensive human intervention, albeit at a cost structure that doesn't obviously translate outside markets with large, affordable gig workforces. Soulgate's approach—leaning into AI, gamification, and virtual interaction—maps more closely to where Tinder, Bumble, and others are already investing product resources.

What this means for operators outside China

Yidui's 16.4% paying conversion rate will turn heads in product and growth teams across the industry. That number alone invites the question: could human facilitation work in Western markets?

The honest answer is probably not at this cost structure or revenue model. Yidui operates in a market where live-streaming gift economies are culturally embedded and labour costs allow for a 193,000-person host network. Replicating that in the UK or US—where minimum wage laws, employment regulations, and cultural expectations around tipping and virtual gifts differ—would require a fundamentally different economic model.

But the broader principle—that users might pay significantly more for facilitated interaction than self-directed swiping—deserves attention. Match Group experimented with video dating features during COVID lockdowns and saw uptick in engagement, but never monetised it directly. Bumble introduced video and voice call prompts but similarly treats them as engagement features rather than revenue drivers. Neither platform charges specifically for facilitated introductions or mediated conversations.

A lighter-touch version might work: imagine a premium tier where trained facilitators offer optional 10-minute icebreaker video calls between matches, charged per session rather than as part of a flat subscription. The economics would differ—probably closer to coaching or concierge services than entertainment—but the hypothesis is testable: users will pay more for help than for access.

The IPO filing doesn't disclose a timeline for the listing, but the financials position Yidui as one of the most profitable dating-adjacent platforms globally by margin. Operating profit margin hit 11.4% for fiscal 2024. Whether investors price it as a dating company, a live-streaming platform, or something in between will determine how seriously Western operators take the model.