Overview

Grouper was a New York social club that turned a blind date into a six-person outing. One member brought two friends, Grouper matched them with another trio, selected a bar, made the reservation, and included the first drink for $20 a person. Founded in 2011 and admitted to Y Combinator's Winter 2012 batch, it expanded from New York to a claimed 25 cities.^[1]

The group format worked. The company reported repeat intent, real revenue, and hundreds of partner bars. The business around it did not. Grouper spread a labor-heavy local marketplace before it had solved cancellations, balanced supply, or repeatable city economics. Then Tinder reduced the same fear of rejection to a mutual swipe, free to the user and nearly free to fulfill.

Grouper cut international operations and staff in June 2014. Database sources mark it closed in October 2016, although no contemporaneous shutdown statement explains the final two years.^[10][13]

Image 1 / 3

Founding Story

Michael Waxman arrived at Grouper with one failed startup behind him. After his freshman year at Yale, he dropped out to build Batiq, a language-learning company inspired by a summer teaching English in Taiwan. The company raised money, hired a team, and failed after about two years. Waxman returned to Yale, learned to program, and spent roughly 40 hours a week building side projects. He later said the earlier failure taught him that a technology founder who could not code lacked basic literacy.^[2]

Grouper came from a more personal break. In 2011, after a rejected YC interview and a breakup, Waxman moved to New York as a newly single graduate who disliked both online dating and the Manhattan bar scene. He built the prototype in one week and launched it on July 6. In a 2013 interview, Waxman described the premise plainly: "modern digital technology can be really isolating."^[2] The product was meant to use software briefly, then get out of the way.

YC partner Garry Tan supplied an early public endorsement in June 2012, sharing the story of Waxman meeting his girlfriend through Grouper and writing that the product "rocks."^[17]

G

Garrytan@garrytan

Michael Waxman, CEO Of Dating Startup Grouper, Found Love On His Own Site http://t.co/wX1OmEfF - Great story! Grouper rocks.

2012-06-13

The earliest founder record is messy. Observer called Jerry Guo and Waxman co-founders at the September 2011 public launch and said the two had met as Yale freshmen. Guo supplied the better creation myth: "We took our $400 travel reimbursement check from YC and used that as seed capital for Grouper."^[3] Guo disappears from later accounts. Waxman subsequently said he applied to YC W2012 alone, while two early employees kept New York operations running. Tom Brown, an MIT engineer working on another YC company, split with his original co-founder and began helping Waxman. Brown joined Grouper shortly before Demo Day.^[2]

The roster therefore evolved rather than fitting one clean founding moment. YC now identifies Waxman and Brown as founders and Challen Hodson and Kristen Badal as members of the founding team.^[1] What stayed constant was the first insight: a blind meeting felt less exposed when both people arrived with friends.

Timeline

• July 6, 2011: Waxman launched a prototype built in one week in New York.^[2]
• September 8, 2011: Grouper launched publicly out of beta at $20 per person; Observer reported hundreds of outings and profitability after 60 days.^[3]
• Winter 2012: Grouper joined Y Combinator; Brown became Waxman's co-founder before Demo Day.^[1]
• September 20, 2012: The service announced a redesign and availability in 12 named US cities.^[4]
• April 30, 2013: Grouper launched an iPhone app in 20 cities, with on-demand outings and a mobile concierge.^[6]
• May 2013: TechCrunch named Tinder and Lulu as a new competitive wave.^[7]
• November 2013: Waxman said Grouper operated in 25 cities and annual revenue exceeded $500,000.^[2]
• November 25, 2013: Funding databases record an undisclosed Series A involving Floodgate; the total amount raised remains unresolved.^[9]
• June 3, 2014: VentureBeat reported international closures and layoffs in the United States and overseas.^[10]
• June 2014: Brown left amid burnout and falling revenue, according to his 2025 account.^[12]
• October 1, 2016: Database sources mark Grouper closed; no contemporaneous announcement was found.^[13]

What They Built

Grouper sold one compact promise: bring two friends and meet another group of three over drinks. The lead participant applied with Facebook, supplied preferences and availability, and named two friends. Grouper combined application data, Facebook information, software matching, and human judgment to select the opposite trio. Users did not browse profiles or message matches. Grouper withheld names and photos, chose the venue, booked a table, and arranged the first round.^[4]

That design removed several awkward moments at once. Nobody had to ask a stranger out directly. Friends supplied social proof and conversation insurance. Prepayment meant the table could leave after one round without negotiating a shared tab. The event could become a date, a hookup, or a new friendship, but Grouper resisted promising any one outcome. Waxman compared the product to a bar rather than a dating site because the company did not expose profiles or permit online conversation.^[2]

The experience only looked simple because staff absorbed the complexity. Matchers assembled compatible trios. A member-experience team handled venue changes, late groups, cancellations, no-shows, and parties that arrived with the wrong number of people. Waxman described software that triaged calls, texts, and email to an on-call team. A contract engineer later documented a Ruby on Rails API, Neo4j, Elasticsearch, asynchronous billing, and a matching engine based on the stable-marriage algorithm.^[2][15]

The April 2013 iPhone app moved booking to the device people carried into the night. It added a concierge, automatic confirmations, Instagram-linked photos, post-event ratings, and a planned option to arrange an outing in under an hour. Bad ratings alerted customer service. TechCrunch reported 400 partner bars and 20 cities at launch.^[6]

The product also extended beyond the table. A free "Groupergram" promotion sent a photographer to capture an outing, turning the night into shareable proof while keeping individual profiles out of the booking flow. Afterward, ratings gave the team a private signal about group fit, venue quality, and bad behavior. Those touches reinforced the club positioning: Grouper sold a produced social occasion, not access to a searchable database. They also widened the operational surface, because photography, moderation, and recovery all depended on the company coordinating another real-world handoff.^[6]

The app made the request faster, not the fulfillment. An under-an-hour Grouper still required two viable trios, a suitable venue, six confirmed people, payment, and a recovery path if any component failed. That distinction became the company's central operating problem at scale.

Market Position

Target Customers

Grouper targeted urban young adults who wanted to meet people without creating a dating profile or enduring a one-on-one blind date. Its strongest early wedge was a lead user who could recruit two friends. That requirement made every booking a referral loop: one customer exposed at least two more people to the product. The group also made the experience easier to try because a bad match did not strand anyone alone.

The service initially skewed toward heterosexual dating, although it arranged same-sex outings and insisted on the broader social-club label. This positioning gave Grouper room to talk about friendship, nightlife, and city discovery. It also obscured the category users were choosing. Brown later described Grouper directly as group dating and evaluated it against Tinder. The practical market was not every offline social interaction; it was low-pressure romantic discovery with friends present.

Market Size

No reliable contemporary market-size estimate specific to curated group dates was found. Grouper's own operating data is more useful. It moved from New York and San Francisco to 12 cities by September 2012, 20 by April 2013, and 25 by November. Waxman said annual revenue exceeded $500,000 in 2013, while the company reported hundreds of thousands of drinks and 400 partner bars.^[2][6]

Those figures proved that people would pay for the format, but not that each city had enough compatible trios at the same time. A six-person event multiplies marketplace constraints. Grouper needed one suitable lead on each side, four available friends, compatible intent, balanced demographics, a shared time window, and a participating venue. A city could look active in aggregate while producing thin inventory for any specific Thursday night.

Competition

Grouper competed on outcome depth. Dating sites offered profiles and messages; Grouper delivered a table with six people. That was valuable, but it also meant Groupon-like local operations attached to a consumer matchmaking product. Bars remained the free substitute.

Tinder changed the axis. It launched in 2012 with free, location-aware discovery and a mutual opt-in before messaging. By May 2013 it reported 50 million matches; by February 2014 it reported 600 million swipes a day.^[8] Its advantage was not a better real-world date. It removed the riskiest step at software cost. A person could signal interest without suffering direct rejection, and a woman could control who gained message access. Grouper addressed the same anxieties by coordinating five other people and a bar.

That made the products structurally unequal. Grouper could create a richer night, but Tinder could supply vastly more attempts. Every Tinder match improved its graph without requiring a reservation. Every Grouper added revenue and operational exposure. Once Tinder's free mutual-like mechanic became the default, Grouper's $20 outing had to justify not merely a better interaction but a more expensive, less available way to reach it.

Grouper also faced a cardinality disadvantage. Tinder needed two users to express interest asynchronously. Grouper needed two acceptable leads plus four friends, and all six had to converge on one place and time. Tightening a safety, age, orientation, location, or availability filter removed a much larger share of feasible Grouper combinations. Human curation could improve an individual match, but it could not manufacture density. The broader the promised geography and the shorter the booking window, the more often the company had to relax fit, spend to recruit the missing side, or disappoint a customer.

Business Model

Grouper charged $20 per attendee for each outing. The price included tax, tip, the first drink, matching, venue selection, reservation, and support. Six attendees implied $120 in gross booking value before the venue's drink cost, payment fees, refunds, and member-experience labor. Grouper also explored membership dues and partner economics with bars, Uber, Airbnb, and other startups, but the fetched evidence does not show a durable second revenue stream.^[2]

That transaction had little room for hidden work. Even before acquisition, a $120 booking had to fund six drinks, payment processing, matching, reservation handling, and live support. The record does not disclose Grouper's venue discount or support cost, so contribution margin cannot be reconstructed. Membership could have improved predictability, but no fetched source establishes that it launched or retained paying members.

Waxman said annual revenue was above $500,000 in November 2013. The company did not disclose gross margin or repeat rate. VentureBeat reported at least 30 employees before the June 2014 layoffs.^[10] A clearly labeled inference shows the pressure: at a hypothetical $100,000 to $140,000 loaded annual cost per employee in New York, payroll alone would have been roughly $3 million to $4.2 million, before offices, marketing, venue operations, and refunds. This is not a reported burn figure, but it shows why $500,000-plus revenue could coexist with a fragile business.

Funding data is too inconsistent for a defensible total. Databases agree on YC, SV Angel, Harrison Metal, Floodgate, Initialized Capital, and an undisclosed November 2013 Series A, but not on amounts.^[9]

Traction

Grouper's earliest signals were unusually concrete. Observer reported that it was profitable after 60 days, had arranged hundreds of outings, and saw 93% of participants say they wanted another. Roughly one-third said participants met again, and demand from women was said to be about twice demand from men. These were company-supplied numbers, not audited cohorts, but they weaken any claim that the concept simply failed to attract users.^[3]

Waxman said Grouper had generated about $30,000 to $40,000 before its successful YC interview and more than $500,000 in annual revenue by late 2013. Expansion was rapid. The company claimed hundreds of thousands of drinks, 400 bars, and 25 cities at its peak.^[2][6]

W

Wxmn@wxmn

Thanks for the interview on @grouper @KateRogersNews! It was a lot of fun. http://t.co/lS9BJtuQfu

2013-08-14

The missing metrics matter more than the headline counts. Grouper never disclosed city-level match fill, repeat booking, cancellation rates, contribution margin, customer acquisition cost, or the share of events that needed human intervention. Brown's later statement that revenue was falling by June 2014 indicates that geographic reach did not translate into stable throughput.^[12]

The sequence is therefore stronger evidence than the totals. Grouper expanded while publishing cumulative cities, bars, and drinks, then retreated while revenue declined. Without cohort retention or local fill rates, the public numbers cannot distinguish durable repeat demand from one-time curiosity amplified across newly opened markets.

Post-Mortem

The six-person transaction tax

Grouper's product insight and business burden were the same thing. Friends lowered the emotional cost of a blind date, but each friend added another point of failure. The company could not complete an order unless two suitable leads each recruited two available friends, accepted the time and place, paid, and appeared. Venue quality then became part of the product even though Grouper did not control the bar.

The failure mechanism surfaced early. Waxman said cancellations rose disproportionately when the company scaled in spring 2012. Each cancellation destroyed revenue, created support work, and disappointed the opposing trio. Grouper's remedy was revealing: it required the cancelling group to apologize through a phone-number-shielding app. Social accountability reduced casual cancellations, but it did not remove the underlying combinatorics.^[2]

Early success could therefore conceal later weakness. A few hundred outings in one dense city could be profitable because the team knew the venues, watched every match, and concentrated demand. The same service in 25 cities required local venue knowledge, balanced cohorts, support coverage, and enough simultaneous intent in each market. Company-wide drink counts aggregated those markets and hid the one measure that mattered: whether a specific trio could receive a good counterpart at the requested time without paid acquisition or manual rescue.

The iPhone app then promised speed. On-demand booking in under an hour increased customer convenience while shortening the company's window to match six schedules and reserve a bar. Software triage helped operators react, but it did not turn the service into a software-margin transaction. Brown's later recollection that a team manually linked people confirms that labor sat in the fulfillment path.^[11]

Expansion diluted the thing each city needed

Grouper expanded from two cities to 25 before public evidence showed repeatable local economics. Waxman celebrated the built-in referral loop because every customer brought two friends. Yet he also said the company was testing paid ads to open markets and correct gender imbalance. Virality supplied exposure, not necessarily the right people in the right city for the same night.

This is the non-obvious structural failure: city count worked against match quality when operations and acquisition were spread before local cohorts were dense. A marketplace for six-person, intent-sensitive outings cannot treat nominal availability as liquidity. VentureBeat reported that Grouper still advertised 25 cities while shutting international operations and pausing London in June 2014.^[10] The footprint had become a lagging marketing claim, not proof that each market cleared.

The layoffs cut the people who held that footprint together. Reported departures included international operations, business development, and community leadership. Retrenchment reduced burn, but it also weakened fulfillment.

Tinder solved rejection before the date

The strongest counterargument is that Grouper failed from operational overreach, not competition. The company had revenue, satisfied users, and a format distinct from swiping. Had it stayed in New York, charged more, or expanded later, it might have built a profitable social club.

That argument explains the speed of decline but not the shrinking strategic space. Tinder solved a prior step in the same job. A mutual like removed direct rejection before two people spoke. It made discovery abundant, private, and free. By early 2014 Tinder reported 600 million swipes a day, while Grouper still had to orchestrate one table at a time.^[8]

Brown's 2025 account closes the causal loop. He said, "Our business wasn't succeeding. Our revenue was going down," and judged that Tinder "solved the mission that we were trying to solve better."^[12] Tinder did not reproduce Grouper's night out. It made that expensive night unnecessary for far more users.

The quiet ending

On June 3, 2014, VentureBeat reported international shutdowns and layoffs. Brown says he left that month, burned out and unable to repeat a recruiting pitch he no longer believed. The company continued in some form until 2016, but the public record goes thin after the retrenchment. Wikipedia marks October 1, 2016 as the closure date. No founder announcement, investor account, acquisition record, or bankruptcy filing was located.^[13]

A 2020 Hacker News commenter who said they had used Grouper remembered it as fun and inferred that matching, venue choice, reservations, and drink payment must have been an operational nightmare. That anecdote is not causal proof, but it matches the founder and employee descriptions of the system.^[14]

The evidence supports a firm conclusion through June 2014 and a cautious one afterward. Grouper's revenue was falling as it closed markets and cut the operators behind them. The precise final trigger in 2016 remains unknown.

The missing ending changes attribution at the margin, not the documented break: falling revenue, market retreat, staff cuts, and a co-founder's exit arrived together in June 2014.

Key Lessons

• A friend loop is not marketplace liquidity. Grouper forced every organizer to introduce two users, but still needed paid acquisition to start cities and correct gender imbalances. Invitations grew awareness; they did not guarantee compatible trios for the same night.

• Faster booking can worsen fulfillment. The iPhone app's under-an-hour promise reduced planning for customers while compressing matching, payment, reservation, and recovery for staff. Product convenience moved work behind the curtain instead of removing it.

• Compete on the job's cheapest decisive step. Grouper delivered a richer offline experience. Tinder won the larger market by removing rejection before contact with mutual opt-in, at a cost structure no six-person concierge could match.

• Expansion should follow density evidence. Grouper reached 25 cities without disclosing city-level fill, repeat, cancellation, or contribution margin. The 2014 retreat suggests geography became a vanity metric before each market could clear reliably.

• A quiet wind-down limits what can be claimed. The evidence establishes falling revenue, layoffs, international closure, and Brown's departure in June 2014. It does not establish the exact financial or governance event that ended the company in 2016.

Together, these lessons point to a narrower rebuild: prove one dense weekly cohort and its contribution margin before treating geographic reach as progress.

Sources

1. Y Combinator: Grouper company profile
2. Mixergy: Michael Waxman interview
3. Observer: Grouper public launch
4. TechCrunch: Grouper expands across US cities
5. Public Radio East/NPR: observed Grouper outing
6. TechCrunch: Grouper's iPhone and on-demand launch
7. TechCrunch: Grouper faces Tinder and Lulu
8. TechCrunch: Tinder reaches 600 million daily swipes
9. Dealroom: Grouper funding and investors
10. VentureBeat: Grouper closes international operations and cuts staff
11. Y Combinator Lightcone: Tom Brown interview
12. Searchable transcript of Tom Brown's YC interview
13. Wikipedia: Grouper social club
14. Hacker News: former-user discussion
15. Brett Halperin: Grouper API engineering workpaper
16. Business Insider: Michael Waxman and Grouper
17. Garry Tan: early public endorsement of Grouper
18. Hacker News: Grouper launches across US cities
19. Hacker News: Michael Waxman explains Grouper's model
20. Michael Waxman: Fox Business interview post