Timeleft: 3 years and 4 failed ideas, then dinners with 5 strangers made €18M a year
Americans now spend 34 minutes a day with friends, down from 47 twenty years ago. Two French founders spent 3 years failing to fix that. Then they tested one last idea with a form, a WhatsApp group and a payment link. The first week made €125. Two years later it was €18 million a year.
Overview
Timeleft puts 5 strangers at one restaurant table for about 3 hours. You sign up, answer questions about yourself, and get matched with people you have never met.
It was started in France in 2023 by Maxime Barbier and Adrien De Oliveira. By 2025 it had about 150,000 people at its dinners every month, in more than 200 cities, about 100 staff, and €18 million in annual revenue.
The trend they are riding
The American Time Use Survey tracks how people spend their day. In 2003 the average American spent 47 minutes a day with friends or at social events. In 2013 it was 43. In 2023 it was 34.
For 15 to 24 year olds the drop is steeper: from about an hour a day to about 35 minutes. In 2023 the U.S. Surgeon General called loneliness an epidemic.
A Data for Progress survey (September 2026) compared Americans today with a 1995 survey and found the same shape:
- Entertaining people at home at least once a year: 88% in 1995, 67% in 2026.
- Attending club meetings: 45% to 26%.
- Going on a picnic: 58% to 39%.
- Americans with five or more close friends: 63% in 1990, 38% in 2021 (American Perspectives Survey).
What went up: internet use, video games, and exercise classes (+14 points). People still want to be in a room with others. They just need a reason and a booking button.
Why people can't fix it on their own
Researchers at the University of Kansas estimated it takes about 50 hours together to go from acquaintance to casual friend, about 90 hours to become friends, and more than 200 hours to become close friends. A meetup once a month never gets you there.
Read any long online discussion about loneliness and the same line keeps coming up: somebody has to be the one who invites. Most people are waiting to be invited. The people who do have friends as adults are usually the ones who organize the dinner, the game night or the potluck, and many of them say they hate doing it.
That is the real product. Timeleft doesn't sell dinner. It sells someone else doing the inviting, every week, so you only have to show up.
So the demand is not in doubt. Millions of people want more friends and do not know how to meet them as adults. The question is what to sell them.
3 years of nothing
Maxime came from MinuteBuzz, a French media company. He and Adrien spent about three years testing ideas for bringing people together, and none of them worked.
They tried matching people on their dreams. They tried matching people on activities. They tried group meetups. In their own telling it took 3 years and 4 pivots to earn their first euro.
His own verdict on the early ideas: "We realized that apart from dreaming, it wasn't going to go very far."
The €125 test
In 2023, out of cash and energy, they gave themselves one last chance and a deadline: 15 days to test a single idea. The idea: dinner with strangers.
They did not build an app. They used three tools anyone can set up in an afternoon:
- Typeform for the sign up and the questions that matched people.
- WhatsApp to tell each group where and when to show up.
- Stripe to take payment up front.
The first week made €125. That is a tiny number, but it answered the only question that mattered: would strangers pay to have dinner with other strangers? They would. And the people who came loved it.
The first customers
The first official Timeleft dinner was in Lisbon on May 3, 2023, with 24 people. Lisbon was a smart first city: full of remote workers and newcomers who had just moved there and knew nobody, the exact people who want a stranger dinner most.
Maxime had run events in Paris before Timeleft, so he already knew how to fill a room. Within 14 weeks Timeleft had more than 17,000 members. The mechanism was the waitlist: every new city showed as "coming soon" until enough people signed up, so each city collected members for weeks before its first dinner, and people shared it to get their city open sooner. Paid ads filled the waitlists; the founders say they grew on ads alone, with no city managers or ambassadors. When it launched in Paris on September 27, 2023, more than 120 people showed up the first night, spread across more than 10 restaurants.
Charging from the very first dinner did two jobs. It brought in money, and it filtered out the people who would sign up and never show.
How it works
- The matching. New members answer questions about themselves, and the answers decide who sits with whom.
- The table. 5 strangers, one restaurant, about 3 hours. The restaurant is a partner, not a cost: Timeleft brings it a full table on a quiet night.
- The price. Pay per dinner or subscribe monthly, starting around $26 in most countries, with discounts for 3-month plans.
- The feedback. 15,000 to 20,000 people leave reviews every week, which feeds the matching.
How they solved the empty-table problem
A dinner with strangers only works if enough strangers show up. A city with 12 members gives you a sad, half-empty table.
Their rule: a new city did not open until 151 people had signed up. Until then, the city showed as "coming soon." That did two things. It created scarcity, so people hurried to join, and it guaranteed full tables from the first night.
How it grew
- December 2023: about €20,000 a month in revenue.
- January 2024: first city outside France, Barcelona, opened with ads alone.
- 2024: they opened roughly one city a day for a year, reaching 320+ cities. No city managers, no ambassadors. Just ads and the 151-signup rule.
- 16 months after launch: €1 million in revenue in a single month.
- 2025: about €18 million a year. They pulled back to around 200 cities, because small cities could not fill tables every week. Their rule became 50+ people a week per city.
The tech stayed simple for a long time. The first 16 months ran on Bubble and FlutterFlow, two no-code tools. Only after that did they spend 8 months moving to custom software, once their CTO said: "I can't touch anything anymore, if I touch something, it breaks."
To keep members coming back, they added new formats: bar meetups, women-only dinners and running groups. The average member went from 1.4 to 2.2 events a month.
The numbers
| Figure | Number |
|---|---|
| Time before first euro | About 3 years, 4 pivots |
| Test that worked | Typeform + WhatsApp + Stripe, a 15-day deadline, 2023 |
| First official dinner | Lisbon, May 3, 2023, 24 people |
| Members after 14 weeks | 17,000+ |
| Paris launch night | Sept 27, 2023: 120+ people, 10+ restaurants |
| First week | €125 |
| December 2023 | About €20,000 a month |
| First €1M month | 16 months after launch |
| Annual revenue, 2025 | About €18 million |
| People per month | About 150,000 |
| Cities | 320+ at the peak, about 200 now |
| Signups needed to open a city | 151 |
| Price | From about $26 a dinner, or monthly |
| Staff | About 100 |
| Outside funding | About $7M Series A, July 2024 (Kima Ventures, FJ Labs, Play Ventures, Redstone), after the dinners were already working; a small earlier round with angels |
Revenue figures are the company's own, from interviews with the CEO; an earlier interview gave €12 million.
Others making money from loneliness
- Camp Social: adult summer camp for women who want to make friends. About $883 for two nights.
- 222: an app that pairs strangers for dinners and outings, matched by AI. Raised $10 million.
- Lonely Girls Club (UK): a community for women looking for friends, 35,000 members.
- Friendship apps overall: about $16 million of U.S. consumer spending in 2025 across more than a dozen apps (Appfigures estimate).
Do what they did: step by step
1. Pick a pain everyone already feels.
Do this: look for a trend with public data behind it. Loneliness has the Time Use Survey and a Surgeon General warning. You do not have to convince anyone it is real.
2. Test with a form, a group chat and a payment link.
Do this: before writing any code, run one week with Typeform (or Google Forms), WhatsApp and Stripe. If nobody pays, you lost a week, not three years.
3. Charge from day one.
Do this: even a small price. It tells you the demand is real and keeps out the people who would never show.
4. Do not open until you can fill the room.
Do this: pick your own "151." Collect signups for a city, a class or an event, show it as "coming soon," and launch only when the number is reached.
5. Grow with ads, one place at a time.
Do this: once one city works, copy it to the next with the same ad and the same threshold. Cut the places that cannot fill tables every week.
6. Stay on no-code until it breaks.
Do this: Timeleft ran on Bubble and FlutterFlow for 16 months while making millions. Rebuild only when the tools stop you, not before.
7. Small version you can start this month.
Do this: run a monthly dinner for 6 strangers in your city, $25 to $40 a seat, booked through a form, confirmed on WhatsApp, paid through Stripe. Partner with one restaurant that wants a full table on a slow night.
The kit: run your first dinner this month
1. The signup form (Typeform or Google Forms, 2 minutes to fill):
- First name and phone number (for WhatsApp)
- Age range: 21 to 29 / 30 to 39 / 40 to 49 / 50+
- Which night works: Wednesday / Thursday
- Pick one: "I'd rather talk about ideas" / "about people" / "about adventures"
- Pick one: "At a party I'm the one who starts conversations" / "who joins them" / "who finds the one person to talk to all night"
- Anything you don't eat?
- Last line: "Seats are $30, paid when we confirm your table."
Seat people of similar age together, and mix the answers to the last two questions so every table has a talker and a listener.
2. The WhatsApp message (send 48 hours before):
"Hi [name], you're in. Wednesday, 7:30 pm, table for 6. You'll get the restaurant name tomorrow at noon. Pay here to hold your seat: [Stripe link]. If you can't make it, tell me by tonight and someone on the waitlist gets your seat."
Then on the day: "Tonight: [restaurant], [address], 7:30. Table under 'Dinner Club'. Five people you haven't met are about to become people you have."
3. What to charge
$25 to $40 a seat for the booking, dinner paid separately to the restaurant. That price is high enough that people show up, low enough to say yes on impulse. Timeleft starts around $26.
4. The restaurant pitch (walk in at 3 pm on a weekday, ask for the owner):
"I run small dinners for people who want to meet new friends. I'll bring you a full table of 6 every Wednesday, people who order drinks and stay 3 hours. All I need is a reserved table and a set menu at a fixed price. Can we try it next Wednesday?"
Restaurants say yes because a slow weeknight is their worst problem, and you're solving it for free.
5. The rule that makes it work
Don't book the restaurant until 6 people have paid. Post the dinner in local Facebook groups, your city's subreddit and Nextdoor as "coming soon," and confirm only when the table is full. That is Timeleft's 151 rule at the size of one table.
The math: 2 tables a week at $30 a seat is $360 a week, about $1,500 a month, from one restaurant and a few hours of messages. Add a second night or a second restaurant and you have a side business. Timeleft did the same thing, then did it in 200 cities.
Want 50 more breakdowns like this?
Real businesses with their revenue, tools, costs and how they grew, plus step-by-step guides to start your own. Free.
Get free accessMore breakdowns
Sources
- Inside Timeleft's journey to connecting 150,000 diners monthly (interview with CEO Maxime Barbier, Substack)
- Timeleft blog: "Back to Basics: Timeleft Launches Its Unique Dinners in Paris" and "Dinner with Strangers: The New Social Revolution Hits Paris"
- Temy, profile of Maxime Barbier
- Startup Spells, "How Timeleft solved the chicken and egg problem"
- HNGRY, "Timeleft creates social dining experiences to cure the loneliness epidemic"
- U.S. Bureau of Labor Statistics, American Time Use Survey, as reported by Axios (2026) and Audacy
- Jeffrey Hall, University of Kansas, "How many hours does it take to make a friend?" (Journal of Social and Personal Relationships, 2018)
- Data for Progress, "Americans' social activity has steeply declined in the past 30 years" (September 2026), citing the 1995 DDB Needham Life Style survey and the American Perspectives Survey
- CB Insights and Nordic9 (Timeleft funding rounds)
- TechCrunch (2026) on friendship apps, Appfigures estimates; Yahoo Lifestyle on Camp Social; Wikipedia, Lonely Girls Club
