TLDR
The friendship app market has reached an inflection point. Bumble BFF generates $0 standalone revenue while Timeleft hit €18M ARR in under two years using a facilitated group dinner model. Several well-funded startups (222, Pie, Clyx) are betting AI-matched group events represent the future. Legacy apps Hey! VINA and Friended have shut down or gone dormant. The market is fragmenting fast — and the data shows who's winning.
- Facilitated group model
- A friendship app design where the platform handles end-to-end logistics: algorithmic matching, venue selection, event timing, and conversation structure. Participants just show up. Timeleft (€18M ARR) and 222 (YC-backed) are the leading examples. Contrasts with self-organized models (Bumble BFF, Meetup) where users arrange their own meetings after matching.
DEFINITION
- Swipe-based matching
- A 1:1 matching mechanism borrowed from dating apps and applied to friendship. Research suggests this model has a structural design problem for friendship: it lacks the biological urgency that drives romantic matching, creates ghosting, and produces conversation without action. Bumble BFF is the largest example; $0 standalone revenue as of Q3 2025.
DEFINITION
The friendship app market has split in two, and the split is getting wider.
On one side: swipe-based apps that applied dating mechanics to friendship. Bumble BFF has $0 in standalone revenue despite being embedded in a platform with nearly a billion dollars in annual revenue. Hey! VINA shut down after a decade. Patook is listed as deadpooled. Friended has a 1.6/5 iOS rating and a website full of placeholder text.
On the other side: facilitated group experiences that remove the coordination burden entirely. Timeleft hit €18M ARR in under two years. Three well-funded startups (222, Pie, Clyx) raised $10-14M each in 2024-2025, all betting on the same model. The market hasn’t died — it’s fragmenting along a clear fault line.
The Incumbents: Scale With Structural Problems
Bumble BFF is the largest friendship product by exposure — but that exposure comes from its integration with a dating app, not standalone appeal. The mechanics are identical to Bumble Dating: swipe, match, message, organize your own meetup. As of Q3 2025, this has produced exactly $0 in standalone revenue. Day-30 retention sits at 10%. Q4 2025 churn across the Bumble platform reached 39.7%.
In September 2025, Bumble relaunched BFF with a Groups tab powered by Geneva (acquired in May 2024), adding community-based discovery alongside 1:1 matching. The pivot implicitly acknowledges that pure swipe-to-friendship wasn’t working.
Meetup has 60 million registered users and approximately 500,000 active ones — a gap that says everything about historical churn. Since Bending Spoons acquired Meetup in January 2024, organizer fees have risen sharply (from $108.99 to $214.79 for a 6-month plan within a single year by some accounts), AI-only customer support has replaced human contact, and Trustpilot sentiment has gone negative. Web traffic declined 18.9% month-over-month as of February 2026. Bending Spoons also acquired Eventbrite in December 2025, which raises questions about how the two platforms get positioned against each other.
Timeleft is the exception that proves the rule. Founded in Paris, it pivoted to its weekly facilitated dinner format in May 2023 — what CEO Maxime Barbier calls “a last-chance pivot.” The model removes every friction point: an algorithm assembles groups of 4-6, selects a partner restaurant, and reveals the venue the morning of the event. Participants know each other’s nationalities, industries, and zodiac signs — but not names or photos. They just show up at 7 PM on Wednesday. That model produced €18M ARR within 20 months across 200+ cities with 150,000 participants per month. The subscription is ~$12.99/month. Food is paid separately at the restaurant.
The New Entrants: All Betting on Groups
Three well-funded startups launched in 2024-2025, and all three chose group-facilitated models over swipe-based matching.
222 (Y Combinator-backed) runs curated events at partner venues — wine bars, comedy clubs — with AI-matched groups assembled from a 30+ question personality quiz that classifies users into 16 types. Events cost $22.22 each. The company raised $12.6M from General Catalyst, Upfront Ventures, NEA, and Y Combinator, currently operating in Los Angeles and New York City.
Pie, founded by Andy Dunn (who built Bonobos), uses ChatGPT-driven personality matching to assemble groups of 6 for events like “Coffee with Strangers.” Its differentiating move: Pie pays event hosts to create and run gatherings, applying a marketplace model to social events. The company reports 130,000+ MAU and 50,000 actively joining events across San Francisco, Chicago, and Austin. It raised an $11.5M Series A.
Clyx combines event discovery with AI-powered compatibility matching. Its “Programs” feature creates recurring multi-week event series where the same groups participate together — building repeated exposure over time. Raised $14M from Blitzscaling Ventures in September 2025. Active in Miami and London with 50,000 ticket-buying users.
None of these companies built a swipe product. That’s a signal.
The Dead and Dying
Hey! VINA shut down permanently in late 2024 after roughly 10 years. The Instagram farewell cited “tens of millions of new friends” but the reality was a platform that raised only $1.59M against a problem that requires either massive distribution or significant capital to crack. Bumble BFF, with its built-in user base, was a competitor VINA couldn’t match.
Patook — once notable for its AI-powered anti-flirting classifier — is listed as deadpooled by Tracxn. The official website domain appears lost to a parking page. The Android app received a maintenance update in March 2026, suggesting solo-developer care at minimum scale.
Friended launched in 2016, shut down, revived during COVID, and soft-relaunched in late 2022. Its iOS rating sits at 1.6/5 from 6 reviews. The website contains Lorem ipsum placeholder text. Last app update: July 2023. No VC funding. Classic chicken-and-egg failure.
What the Data Shows
The pattern across the entire landscape is consistent enough to call it a structural finding rather than coincidence.
Facilitated group experiences produce better unit economics than swipe-to-match. The transaction is clear: users pay for a guaranteed social experience, not the possibility of one. Timeleft’s €18M ARR against Bumble BFF’s $0 is the starkest illustration, but Pie, 222, and Clyx are all building on the same insight.
Swipe-based friendship apps have a structural design problem that better execution doesn’t fix. Users describe ghosting, conversations that go nowhere, and “swipe fatigue that feels performative.” The issue isn’t the product teams — it’s that friendship lacks the biological urgency that motivates romantic matching. Bumble BFF’s addition of Geneva-powered groups is an implicit acknowledgment of this.
Bootstrapped friendship apps consistently fail. Friended (~8K users), Patook (~70K peak), Hey! VINA ($1.59M raised) all stalled or shut down. The minimum viable funding for a standalone friendship app appears to be in the $7-15M range based on the new entrants who are gaining traction.
AI-powered group matching is the new standard. Every well-funded entrant from 2024-2026 uses AI for personality scoring and group assembly. The evolution mirrors what happened in dating a decade ago — and suggests the swipe era for friendship is ending.
What This Means for Adults Looking for Apps Today
If you’re an adult trying to make friends through an app in 2026, the practical answer is to look at platforms that remove the coordination burden from you. Timeleft’s weekly Wednesday dinners, 222’s curated venue events, Pie’s hosted gatherings, Clyx’s recurring programs — these put you in a room with matched people without requiring either side to initiate.
The swipe products still exist and some people find them useful. But the retention data (10% Day-30 for Bumble BFF) tells you what most people’s experience looks like. The facilitated group model is harder to build, requires more operational infrastructure, and costs more to run — which is probably why the winners raised $10M+ before launching. It’s also the model producing the only meaningful revenue in the space.
We built Threvi because the data on this was hard to ignore. The platforms winning right now all share the same mechanism: recurring groups, low coordination overhead, and repeated contact with the same people. That’s what we’re building.
Q&A
Is the friendship app market growing or declining?
The market is fragmenting rather than uniformly growing or declining. Swipe-based apps are struggling: Bumble BFF has $0 standalone revenue, Hey! VINA shut down in 2024, Patook is listed as deadpooled. Group-facilitated apps are thriving: Timeleft hit €18M ARR in under 2 years. New entrants with $10-14M in funding (222, Pie, Clyx) are all betting on AI-matched group experiences. The money is clearly following the group model.
Q&A
Why are swipe-based friendship apps failing commercially?
Three structural reasons. First, friendship lacks the biological urgency of romantic matching — people swipe without the same motivation to convert matches to meetings. Second, the 1:1 coordination problem: someone always has to keep initiating, and inertia kills most matches. Third, group formats produce better outcomes for the same price, making swipe-only products hard to justify paying for. Bumble BFF's $0 standalone revenue is the clearest evidence.
Q&A
Which friendship apps are growing in 2026?
Timeleft is the standout (€18M ARR, 3M+ users, 200+ cities). Among newer entrants, 222, Pie, and Clyx each raised $10-14M in 2024-2025 — a signal of investor conviction in the facilitated group model. Peanut (5M+ users, profitable) continues growing in its niche. Meet5 has 777,000+ US downloads targeting adults 40+.
Ready to meet your group?
Frequently asked