Daniel sent us this one — and it's a question that's been rattling around in his head since his grad school days in London. There was this Indian vegetarian buffet on Chapel Market called Veg Buffet. Every surface covered in vegan propaganda, kept prices deliberately low, and he ended up eating there five days a week for the better part of a year. Knew the staff by name. And it got him thinking: why don't we have a subscription model for restaurants? You pay a monthly fee like a gym membership, and you just show up and eat whenever you want. Professionally run kitchen, no grocery shopping, no dishes. Has anyone actually tried this, what's it called, what would it take to get one running, and how much would it cost? It's a genuinely interesting question because it sits at this weird intersection of economics, psychology, and the history of how we've organized eating outside the home.
The thing is, people have tried this. The concept even has a name — it's usually called a subscription restaurant, a meal membership, or sometimes a food club. But before I get into the modern attempts, there's a deeper history here that most people don't know about, and it actually explains why this model shows up in some places and completely fails in others.
I'm guessing the deeper history isn't just "nobody thought of it until Silicon Valley decided to disrupt lunch.
No, it goes back way further. The subscription eating model essentially existed before restaurants as we know them even emerged. In eighteenth-century Paris, before the modern restaurant took off, there were tables d'hôte — literally "host's tables" — where you'd pay a fixed price and sit down at a communal table for whatever the host was serving that day. It wasn't quite a subscription, but it was the precursor to the idea that you don't order à la carte, you just show up and eat what's prepared.
Fixed menu, fixed price, communal seating. That's basically the Veg Buffet model minus the monthly billing.
And the table d'hôte model persisted alongside restaurants for centuries. But the actual subscription version — pay monthly, eat whenever — that emerged in a few specific contexts. The most successful historical example is probably the boarding house. In the nineteenth and early twentieth centuries, especially in American and British cities, boarding houses typically included meals in the rent. You paid weekly or monthly, and you showed up for dinner at a set time. It wasn't a restaurant per se, but it was functionally a meal subscription attached to housing.
Which collapsed as a model when apartments with private kitchens became the norm. The kitchen killed the boarding house.
And that's actually the first thing to understand about why subscription restaurants are hard: they're competing with the private kitchen, which is one of the most subsidized pieces of infrastructure in modern life. Your kitchen is already paid for in your rent or mortgage. The incremental cost of cooking at home is just ingredients. A subscription restaurant has to charge enough to cover ingredients, labor, rent, utilities, and some margin — and make that competitive with what you'd spend on groceries plus the value of your time.
The question isn't "why doesn't this exist" — it's "what conditions make the math work.
The math works best in institutional settings where people are already paying for meals as part of a bundle. College meal plans are the obvious example. You pay per semester, you swipe your card, you eat. The university captures economies of scale, predictable demand, and a captive audience. Military mess halls, corporate cafeterias, hospital canteens — same model, different context.
Those aren't exactly what Daniel's describing. He's talking about a restaurant you choose to join, not one you're assigned to because you're a freshman or a soldier or an employee.
That's where it gets interesting, because the voluntary subscription restaurant has been tried multiple times in the past decade or so, and the track record is... let's call it instructive.
That's the polite version of "most of them failed.
Let me give you some specific examples. There was a startup called MealPass that launched in 2016, backed by a venture capital firm. The model was simple: you paid a monthly fee, somewhere around a hundred to a hundred and twenty dollars, and you got one lunch per day from a rotating selection of restaurants in your area. You couldn't order whatever you wanted — each day there was a curated set of options. The idea was to aggregate demand and get bulk pricing from restaurants.
It was a middleman, not a restaurant itself.
And it worked, sort of, for a while. They got tens of thousands of subscribers in a few cities. But the unit economics were brutal. The restaurants weren't making much on each meal, subscribers wanted variety, and the logistics of coordinating with dozens of independent kitchens was a nightmare. MealPass got acquired by a larger company and eventually shut down.
The middleman problem. You're adding a layer of cost without adding a layer of value that customers will pay for.
Then there's the model where the subscription restaurant actually owns the kitchen. The most prominent recent attempt was a company called EAT Club, which started as a corporate lunch delivery service and eventually pivoted to something closer to a subscription model. But again, they were delivering to offices, not operating a physical restaurant you could walk into.
The closest thing to what Daniel's describing — a physical location you join and visit — that's still rare.
It is, but it's not nonexistent. There's a handful of independent experiments. There was a place in San Francisco called The Perennial that tried a subscription model briefly — pay a monthly fee, get a certain number of meals, plus discounts on additional guests. It didn't last. There's a more successful example in Japan, where subscription-based izakayas have been a thing for decades. You pay a monthly fee, you get a certain number of visits, and the food is simple, high-volume stuff. But even those tend to be more like a membership club with discounted meals rather than true all-you-can-eat on a subscription.
What about the actual all-you-can-eat buffet model that Veg Buffet used? Why doesn't that just convert to a monthly billing cycle?
Because the economics of an all-you-can-eat buffet depend on a delicate balance. You're betting that the average customer eats a certain amount, and you price per visit accordingly. If you switch to a monthly subscription, you lose the ability to price per visit, and you attract exactly the customers you don't want — the ones who will eat there constantly and maximize their value. It's adverse selection. The gym membership comparison Daniel made is actually perfect, but not in the way he meant.
Gyms make money because most members don't show up. A restaurant subscription attracts people who definitely will show up.
They'll show up hungry. A gym can sell a thousand memberships and only have space for a hundred people at a time because they know eighty percent won't come on any given day. A restaurant subscription doesn't have that luxury. If you sell too many memberships, you get overcrowded at peak times. If you sell too few, you can't cover fixed costs.
You need a model where the demand is naturally staggered, or where the food cost per person is low enough that even heavy users don't sink you.
This is exactly where Veg Buffet was clever, whether they knew it or not. Vegetarian Indian food — rice, lentils, chickpeas, vegetables in sauce — the ingredients are incredibly cheap. The most expensive thing in that kitchen was probably the spices, and even those are cheap in bulk. They could afford to let people eat a lot because the marginal cost of an extra plate was maybe a pound or two.
Plus, Chapel Market in Islington — the rent situation matters enormously. What was the rent on a place like that back when Daniel was there?
I don't have the exact figure for that specific location, but Chapel Market is an outdoor market street in Islington. It's not a high-end retail corridor. The rents for a small restaurant space there, especially in the early 2000s or 2010s, would have been dramatically lower than, say, Upper Street a block away. Low rent plus low ingredient costs equals a business model that can tolerate heavy eaters.
Which brings us to the question of what it would actually take to run one of these. And I think the answer starts with a number Daniel might find painful.
Let's run the numbers. If you're opening a subscription restaurant in a city like Jerusalem — or Dublin, or London — you need to cover rent, labor, ingredients, utilities, insurance, and some margin for the operator. For a small space, rent might be two to three thousand dollars a month. Labor for a single cook and maybe one front-of-house person, even at minimum wage, is going to run you four to five thousand a month with taxes and benefits. Ingredients for, say, thirty people eating one to two meals a day, seven days a week... even with cheap vegetarian staples, you're looking at maybe three to four thousand a month. Utilities, insurance, maintenance — another thousand. So all in, you're at roughly ten to thirteen thousand dollars a month in operating costs for a very lean operation.
That's before the owner pays themselves anything meaningful.
So if you want to break even with thirty members, each member needs to pay around three hundred fifty to four hundred fifty dollars a month. With fifty members, you're down to two hundred to two seventy a month. With a hundred, you might get it under a hundred fifty.
A hundred members eating one to two meals a day is a lot of volume for a small kitchen and a small dining room. You'd need to stagger visits, or you'd need a bigger space, which means higher rent, which pushes the numbers back up. It's a spiral.
This is why the most durable examples of this model aren't standalone businesses — they're attached to something else that subsidizes them or spreads the risk. The college dining hall is subsidized by tuition. The corporate cafeteria is a perk that reduces turnover and increases productivity. The boarding house bundles meals with rent. Even the Israeli kibbutz dining hall, which was functionally a meal subscription for members, was part of a larger communal economic structure.
The standalone subscription restaurant is trying to do something that usually requires an institutional anchor, without the anchor.
— and this is where I think Daniel's instinct is actually right — there's a cultural shift happening that might change the math. More people are living alone, especially in cities. In Stockholm, something like fifty percent of households are single-person. In Manhattan, it's similar. Single people cook less. They eat out more. They're more likely to value the social aspect of a communal dining space.
They're more likely to see the kitchen as wasted space in an apartment where square footage is expensive.
If your kitchen represents a quarter of your studio apartment that costs two thousand a month, you're paying five hundred dollars a month for that kitchen before you even buy groceries. At that point, a meal subscription at three hundred a month starts to look like a bargain — you can ditch the kitchen, get a smaller place, and eat better.
Though most landlords aren't going to let you remove the kitchen. The building code assumes a kitchen exists.
Sure, but you can stop using it. And there's a growing number of people who effectively do. There's a term for this in urban studies: the "post-kitchen household." These are people who eat out for almost every meal, or rely on delivery, or graze on prepared foods. They use the kitchen for coffee and maybe toast. A subscription restaurant is basically the logical endpoint of that trend.
What are the actual attempts that are working right now? Not the venture-backed startups that flamed out, but real businesses that someone could visit?
The most interesting current example is probably in India, where the traditional "tiffin" delivery system has evolved into something close to a subscription model. In Mumbai, there are companies that deliver home-cooked meals to your door daily for a monthly fee. It's not a restaurant you visit, but it's the same concept — professional kitchen, monthly billing, no cooking. The price is remarkably low, sometimes as little as fifty to eighty dollars a month for a daily lunch, because they're operating at enormous scale with very low labor costs.
The dabbawala system plus a billing cycle.
And there are similar experiments in other countries. In South Korea, there are "dosirak" subscription services that deliver lunch boxes. In parts of Europe, there are community-supported kitchens — basically the CSA model applied to prepared meals instead of raw produce. You subscribe to a kitchen, they make a set number of meals per week, you pick them up or eat on site.
Community-supported kitchen. That's the closest label I've heard to what Daniel's describing. But it's still not the "walk in and eat whenever" model.
Because the "walk in and eat whenever" part is the hardest piece. It requires the kitchen to predict demand, prepare enough food but not too much, and manage peak crowding. Most successful models solve this by constraining one of those variables. College meal plans constrain the customer base to a known population. Tiffin services constrain the menu and delivery time. Corporate cafeterias constrain the hours. The completely open "come whenever, eat whatever" subscription is the hardest version of the problem.
Which is why Veg Buffet worked as a pay-per-visit buffet but might not have worked as a subscription. The pay-per-visit model naturally throttles demand — you think twice before going twice in one day if each visit costs you.
Though Daniel was going five days a week. At that frequency, a subscription probably would have saved him money. And this gets to the core tension: the subscription model appeals most to the heaviest users, who are also the least profitable customers. The business needs a mix of light and heavy users to work, just like a gym.
What would it take to get one running in a neighborhood like ours? Let's actually answer the practical part of the question.
I think you'd need five things. First, a low-rent space. Not necessarily a prime retail location — it could be a basement, a converted apartment, something off the main street. Second, a cuisine with low ingredient costs. Vegetarian Indian, as Daniel experienced, is basically ideal. Rice and beans and vegetables. Third, a constrained menu. Not a full restaurant menu — think two or three options per day, maybe one constant staple. Fourth, a member cap. You don't sell unlimited memberships. You sell exactly as many as your kitchen and dining room can handle at peak times, and you run a waiting list. Fifth, a community that actually wants to eat together. The social aspect has to be part of the value proposition, because if it's just about cheap food, people will compare it to cooking at home and it'll lose.
The social aspect is doing a lot of work there. You're basically describing a modern boarding house without the boarding. Or a secular version of a community kitchen.
That framing might be the key to making it work. If you call it a "subscription restaurant," people compare it to a restaurant and expect restaurant-level variety and service. If you call it a "community kitchen" or a "supper club," you set different expectations. You're not selling unlimited choice. You're selling membership in a thing.
"Like a gym membership." That was Daniel's phrase, and I think it reveals something about how people think about this. The gym membership is a commitment device. You pay to create an obligation to yourself. But a restaurant subscription would be the opposite — you're paying to remove an obligation. You're paying to never have to think about what's for dinner.
Which is actually a compelling value proposition for a certain kind of person. The question is whether there are enough of those people in one neighborhood to sustain a business.
Let's talk about what Daniel would actually pay for this, because he asked about cost. You ran the numbers for operating costs. What would the membership pricing look like in a real scenario?
If we're talking about Jerusalem, or a comparable city, and a small community kitchen with maybe thirty to fifty members... I think you're looking at two hundred fifty to four hundred dollars per person per month for one meal a day, maybe four to six hundred for two meals a day. That's assuming vegetarian food, a modest space, and a lean operation where the cook might also be the owner.
For context, what does the average person spend on groceries per month?
In Israel, a single person's monthly grocery bill is probably in the range of two hundred fifty to four hundred dollars, depending on how much they cook and what they buy. So the subscription is comparable to groceries, maybe a bit more, but you're also getting the labor of cooking and cleaning included. And you're getting the social experience.
The math is plausible for a single person who values convenience and community. It's not plausible for a family of four.
No, families have their own economies of scale. Cooking for four is barely more work than cooking for one, but the ingredient cost scales more slowly. A family of four spending a thousand dollars a month on groceries would need a subscription to cost two hundred fifty per person to break even, and that's almost certainly below what the restaurant could charge.
Which suggests the target market is single urban professionals, students, and maybe empty-nesters — people who are cooking for one and don't want to.
That's exactly the demographic that's growing. The number of single-person households has been rising for decades in almost every developed country. In the European Union, it's over thirty percent of all households. In some Nordic countries, it's over forty percent. These are people who might prefer a meal subscription to a private kitchen.
The concept isn't crazy. The economics are tight but not impossible. The real barrier isn't the idea — it's execution, location, and community-building.
There's one more barrier that I think is underexplored: regulation. Restaurants are heavily regulated. Health inspections, food safety certifications, zoning laws, liquor licensing if you want to serve alcohol. A subscription model might not fit neatly into existing regulatory categories. Is it a restaurant? A private club? A catering operation? The answer affects everything from your tax situation to your inspection schedule.
A private club might actually be the smartest regulatory hack. If you structure it as a members-only club, you might avoid some of the requirements that apply to public restaurants. But then you're limited in who you can serve, and you might need a liquor license anyway if you want to serve beer or wine.
In some jurisdictions, private clubs have their own regulatory headaches. It's not a free pass. But it's an interesting thought — the supper club model, where it's technically not open to the public, has been used by underground restaurants and pop-ups for years.
Daniel also asked what this concept is called. We've thrown around a few terms. Is there an established name?
Not really, which is telling. The fact that there's no widely recognized term for "restaurant you subscribe to" suggests the concept hasn't stabilized into a recognizable category. You see "meal subscription," "food club," "community kitchen," "supper club," "subscription canteen." The closest established category is probably "community-supported kitchen," which borrows the CSA model from agriculture. But even that's obscure enough that most people wouldn't know what you meant.
The naming problem is a real business problem. If customers can't easily describe what you are, they can't easily tell their friends about you.
They can't easily search for you. Though I'll note that "Veg Buffet" didn't need a fancy category name. It was just an Indian buffet that happened to be cheap enough and good enough that a grad student could eat there every day. The category didn't matter — the execution did.
That's probably the most important point. Daniel's question is "has anyone done this" and "what would it take," but the real answer is that the concept works when someone just does it well in a specific place for a specific community. The business model matters, but the food and the vibe matter more.
That brings us back to Veg Buffet itself. I tried to find out if it's still open. The search results are inconclusive — there's no clear current listing for a restaurant by that exact name on Chapel Market. It may have closed, or it may have changed names. Chapel Market itself is still very much there, still a working market street with food stalls and shops. But the specific institution Daniel remembers may not have survived.
Which is almost certainly a rent story. If the lease came up for renewal and the landlord saw an opportunity to raise the rent, a low-margin buffet can't absorb that. The economics that made it work in the first place are fragile.
That's the thing about these institutions — they exist in a specific window of time, with a specific landlord, a specific rent, a specific operator who's willing to work insane hours for thin margins. When any of those variables change, the whole thing can collapse.
If someone wanted to start one of these today, the most important decision isn't the menu or the pricing or the name. It's finding a space with a landlord who won't kill you on the rent in year three.
Or buying the building. If you can own the real estate, you remove the biggest variable cost. But that requires capital that most people starting a community kitchen don't have.
Which is why these things tend to emerge in spaces that are already owned by someone with a mission. A church basement. A community center. A co-op building. The real estate subsidy is the invisible ingredient.
That's actually how a lot of the most successful community eating models work. The Sikh langar — the free community kitchen in gurdwaras — is the most impressive example. It feeds anyone who shows up, regardless of religion, for free. It works because the community donates the food and labor, and the space is already owned by the religious institution. It's not a subscription model, but it's proof that communal eating at scale is possible when the real estate and labor are subsidized.
The langar is basically the platonic ideal of what Daniel's describing, minus the billing. Everyone chips in, everyone eats. The subsidy is volunteer labor and donated ingredients rather than membership fees.
There are secular versions emerging. In the UK, there's a growing network of "community cafes" and "pay-what-you-can" restaurants that operate on a hybrid model — some people pay full price, some pay less, some volunteer in exchange for meals. It's not a clean subscription model, but it's solving the same problem: making prepared food accessible and communal without the full restaurant markup.
The answer to "has anyone done this" is: yes, in many forms, but rarely as a pure subscription restaurant. The successful versions tend to be either institutionally anchored like college meal plans, culturally embedded like the tiffin system, or subsidized by something other than customer payments, like the langar or community cafes.
The answer to "what would it take" is: low rent, cheap ingredients, a constrained menu, a member cap, and a community that values the social experience enough to pay for it. Plus, probably, an operator who's willing to work very hard for not very much money.
The operator problem is real. Running a kitchen is hard work. Running a kitchen where your customers have a sense of ownership because they're paying monthly is even harder. They'll have opinions. They'll want the menu to change. They'll show up at weird hours. The customer-is-always-right dynamic gets complicated when the customer is also a member.
Though that cuts both ways. Members who feel ownership might also be more forgiving, more willing to help out, more invested in the place's success. A well-run community kitchen can develop a culture where people bus their own tables, don't waste food, and treat the staff well because it's "their" place.
Like a co-op more than a restaurant.
And there are food co-ops that operate something close to this model. The Park Slope Food Coop in Brooklyn has been running for decades — members work shifts in exchange for access to cheap groceries. It's not prepared food, but the membership model is proven. Extending it to a prepared-food kitchen isn't a huge conceptual leap.
It's a huge operational leap, though. Groceries are shelf-stable or refrigerated. Prepared food has to be cooked and served within hours. The logistics are completely different.
But people have solved harder logistics problems. The dabbawalas of Mumbai deliver hundreds of thousands of lunches a day with an error rate that's basically zero, using a system that's been refined over more than a century. The problem isn't that it's impossible — it's that it's hard to make the economics work without the cultural infrastructure that supports those existing systems.
Let's bring this back to Daniel's specific situation. He's in Jerusalem. He wants to know if someone could start one of these in his neighborhood. What would the pitch look like?
I think you'd start with a pilot. Find twenty people who are interested, charge them a flat monthly fee — maybe three hundred dollars for one meal a day, weekdays only to keep it simple. Rent a small kitchen space, or partner with an existing restaurant that's closed during the day. Serve two or three rotating dishes, vegetarian, heavy on rice and legumes. Run it for three months and see if the numbers work.
The "partner with an existing restaurant" idea is smart. A lot of restaurants have idle kitchen capacity during off-hours. A lunch-only subscription service operating out of a dinner-only restaurant's kitchen is basically free rent.
It solves the equipment problem. The kitchen already has the stoves and the walk-in and the prep tables. You're just using them when the main business isn't.
There's a term for this in the restaurant industry — ghost kitchens, dark kitchens, commissary kitchens. The idea of sharing kitchen space is well established. Applying it to a subscription model is novel but not crazy.
It's actually being done in some places. There's a company in London called Karma that started as a food-waste app and evolved into something closer to a meal subscription. There are "community fridges" in neighborhoods across the UK and the US where people share surplus food. The pieces are all there. Someone just needs to assemble them into a subscription kitchen.
The question is whether the person assembling them can make a living doing it. And that's where I think the answer is: probably not at first. This feels like something that starts as a side project, or a labor of love, and maybe eventually becomes sustainable if the community grows enough.
Which is basically how Veg Buffet operated. The owners kept prices deliberately low, which means they weren't maximizing profit. They were running a business, but they were also running a mission — the vegan propaganda everywhere wasn't exactly a profit center. They wanted to make vegetarian food accessible, and the business model served the mission.
The vegan propaganda in the bathroom is such a specific detail. You're sitting there after your third plate of curry, and the wall is telling you about lower cardiovascular disease rates. It's like the restaurant equivalent of those motivational posters in a gym — you're already here, but here's why you should keep coming.
It's brilliant, honestly. Every surface reinforcing the value proposition. You're not just eating cheap curry. You're making an ethical choice. You're improving your health. You're part of something.
The glockenspiel of corporate approachability, except it's the bathroom wall of vegetarian evangelism.
I don't know what that means, but I'm going to nod and move on.
That's fair. So to summarize what we've actually answered: the concept exists but doesn't have a stable name. The closest labels are community-supported kitchen, meal subscription, or food club. The economics are tight — you need low rent, cheap ingredients, and a member cap. The most successful versions are institutionally anchored. The pure standalone subscription restaurant is rare and most attempts have failed, but the underlying trends — more single-person households, more people eating out, more interest in communal dining — suggest the window is opening, not closing.
If someone wanted to start one, the smartest approach is probably a pilot with an existing kitchen, a constrained menu, a small committed group, and a clear understanding that this is a community enterprise, not a get-rich-quick scheme.
The gym membership analogy is the right instinct, but the wrong economics. A gym wants members who don't show up. A subscription restaurant needs members who do show up, just not all at the same time.
That's probably the single sentence that captures the whole challenge.
The question of cost — Daniel's last specific ask — we landed at roughly two fifty to four hundred a month for one meal a day in a city like Jerusalem, assuming a lean vegetarian operation. More for two meals. Less if you can find a subsidized space or share a kitchen. That's comparable to groceries for a single person, and you're trading the cost of cooking time and kitchen cleanup for the premium over raw ingredients.
For the right person, in the right neighborhood, that trade makes a lot of sense. The fact that it doesn't exist everywhere doesn't mean it can't work. It means it's hard, and the people who pull it off are usually doing it for reasons beyond pure profit.
Like covering every surface in vegan propaganda.
Exactly like that.
And now: Hilbert's daily fun fact.
Hilbert: In the 1720s, a Tibetan monk named Ngawang Tashi produced a mathematical manuscript that used a unique system of color-coded numerals — red for positive quantities, black for negative, and gold leaf for zero — making it possibly the only known mathematical text in history where the number zero was literally gilded.
Gold leaf for zero. That's a level of respect for nothing I was not prepared for.
honestly kind of beautiful.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop, and to everyone listening. If you enjoyed this episode, leave us a review wherever you get your podcasts — it helps more people find the show. We'll be back soon with another prompt.
