The points the article make come close to my gripe with ghost kitchens but don't quite cover it:
they feel like scams and when I've accidentally ordered from a ghost kitchen it was by design a terrible experience.
I'm talking like, you order a 15$ main that is called "creamy pasta with prosicutto" and when it shows up its buttered spaghetti with a couple stamp-sized bits of ham. Ordering from actual restaurants come with some of the downsides the article assigns to ghost kitchens, like cold food and weird presentation, but ghost kitchens never seemed to reach the bar of "food someone would actually order, even if it was teleported to them instantly".
I mean they basically are drive by scams. They just flood the market with a million listings for the same kitchen, use some stock photos (AI generated now). And if you get bad reviews or food poisoning complaints you delete the business and list up 5 more.
>>"food someone would actually order, even if it was teleported to them instantly".
The article states
>>Quality control became impossible. Shared kitchen facilities meant that one staff member prepared food for multiple brands simultaneously. No ownership. No accountability. Just assembly-line cooking with zero connection to customers.
I'm not sure if it was impossible or if management never actually prioritized it, not bothering to understand what an actual customer would want. How much of it is the stupid management assumption that they can "just make a dish generally meeting description X on the menu" and deliver that and it'll be ok? «— Real question, did mgt fail at the product specification level, or was QC just as a practical matter, impossible?
On the economics, it really seems 30% for delivery is insane. It seems that same 30% might exceed the cost of the physical restaurant. And when it adds a 15-45min delay while homogenizing and cooling the meal, it seems an impossible problem. Maybe if the 30% transported it instantly and losslessly...
Probably good this soulless idea will die. Too bad so much perfectly good capital was squandered on it instead of better ideas
It really seems like it should be possible, but you have to put in the effort to develop recipes, buy minimum quality ingredients, and train the staff. Old school diners, especially Greek diners in the NYC area, used to be famous for their wide-ranging menus—burgers, spaghetti, spanakopita, chopped liver, etc.—and the food was generally pretty good. Cheesecake Factory has built something similar on a national level, and workplace cafeterias often aren't bad either, certainly not at the level of a ghost kitchen.
I think tech founders often underestimate what it takes to build a food business and what the margins are like and then start to cut corners to make the business viable.
> Just assembly-line cooking with zero connection to customers.
And this is how it works in many (not all) American airports. Local restaurants put their brands on the signs, but the food is prepared by probationary employees of Acme Baggage Displacement And Cafeteria Management Corp.
It would be really nice to have a tag on HN to filter out LLM-generated, or at least partly AI-generated content like this.
If an article makes it to the front page despite being AI-generated it probably has some interesting points or ideas, but it's unfortunate that people seeem to choose the speed and style of LLM writing over the individual style and choice that made the writing of yesteryear more interesting and unique.
I was thinking about commenting the same thing. It had an awful lot of paragraphs that ended in a list of three sentence fragments, usually noun phrases, sometimes negated ones. Was that what tipped you off?
If you believe that, it’s best to flag the submission and move on rather than pollute the comments. This is the equivalent of posting “this is a badly written article”. It doesn’t add any value.
You don't speak for us. If you are going to demand supporting evidence for obvious statements, then you can present supporting evidence for your spurious claims about value.
If an article is badly written or AI-generated, there's value in that being pointed out in the comments. It can save people wasting time, and ideally, discourage people from posting low quality content. That's a large part of the point of a site like this.
> No dining room. No servers. No storefront. No customers walking through the door. Just a kitchen.
> No ownership. No accountability. Just assembly-line cooking with zero connection to customers.
> No loyal regulars. No servers to smooth over problems. Just angry reviews that destroyed virtual brands forever.
Pretty common pattern these days.
That, plus the hashtags at the end (unless Substack uses those and I was unaware of it), plus the fact that we know he's using AI in some capacity because of the feature images - it's a reasonable conclusion to draw.
Anecdotal: Due to lifestyle factors, in my family we use Ubereats often, probably 4-5 orders per week on average.
I have never placed an order from a restaurant that I don't actually know beforehand. Maybe I haven't been there in person, but I know the name of the restaurant and I remember where is it and maybe someone I know ate there.
The only time I go completely blind is if I'm traveling. But that's like 1-2% of my total orders. And, in that case I usually stick to known chains or maybe I ask what's a good restaurant around here, then find it in the app.
I suspect it's not so uncommon to order from restaurants one already knows, rather than taking risks?
If that's the case, then ghost kitchens are going to have a very hard time getting their slice of the market
Eating out in US or any developed western civilization is so expensive that everyone in those civilizations think cooking is a basic skill.
Eating out in US every meal would net $40-$60 a day. Multiplying by 30 days yields an ungodly amount of money. If you doordash it, then just multiply it by 2-3x.
I don't understand why there is no good cheap option. In Japan, I could go to a low-tier shop like Sukiya. It would cost $2-3; the food is decent with taste, and it fills my stomach. In US, $2-3 would be the min amount of tips.
Food that travels well requires different recipes, different ingredients, different packaging.
This doesn't ring true, given the popularity of DoorDash and the fact that most of those restaurant menus are in no way optimized for delivery.
Having lower-quality multitasking staff serving multiple restaurants from the same prep and equipment seems like an obvious way to lose quality. But simply streamlining the process of getting a menu onto delivery? That seems like a solved problem.
> This doesn't ring true, given the popularity of DoorDash and the fact that most of those restaurant menus are in no way optimized for delivery.
People have really funny ideas about restaurants. Somebody once left an online review of my family's establishment complaining that the hot chicken that was supposed to be on their cold to-go salad was in a separate container. They asserted that it was a "trick" to keep the chicken warm and moist, as though it would have been better to let the hot poultry heat their salad in the same container until it was lukewarm meat on top of wilted greens. Every day I wake up and mourn the IQ point that I lost reading it.
So some places are optimizing their fries for delivery.
I've also noticed some restaurants are better at adapting the packaging, like punching out ventilation so fried products don't steam themselves in transit. Lawrence Seafood (which rules) did that for a side of tempura we got this weekend.
But I agree in large part. I wouldn't order fried chicken delivered via door dash in any event. People doing that are optimizing for something other than quality.
Ten years ago (I moved since), there was a ghost kitchen in the Bay Area that was great. I wish I could remember the name. But it produced meals that were extremely well packaged, but designed to be reheated at home (it took like 15 min max). It was great because the meals could be pre-ordered and delivered by the time you got home. You could tell the recipes were all tweaked to fit the specifics of delivery and reheating.
This was an example of a well functioning ghost kitchen. I don’t know how profitable they were, but it was very convenient. There are a lot of downsides to this approach, like pre ordering, reheating, and limited menu, but it was a very different approach to current ghost kitchens around me now or DoorDash from a local restaurant.
Every greasy spoon I frequent now has all sorts of packaging/to-go options, they're self-branding the cups and boxes, they have separate queues for pick-up/delivery orders, they have cubbies for quick pick-up, whatever -- they're all seemingly optimizing for pick-up/online stuff.
Yes, I am sure about this claim that I am making. Are there restaurants that do optimize for delivery? Certainly. But DoorDash covers most restaurants in my area (and it's a big area --- Chicagoland) and most of those menus are identical to the in-person menu.
> This doesn't ring true, given the popularity of DoorDash and the fact that most of those restaurant menus are in no way optimized for delivery.
People are dumb.... but only for so long. I've been burnt so many times by delivery I've gone back to mostly ordering pizza and picking it up in person.
Asimov was fond of the trope that the future was starved of protein, and even america had become a land where people had to eat communally, and eat significant amounts of manufactured "zymoveal" protein, because real meat was scarce under population/land pressure. It was clear that "people didn't like it"
Oddly, "cafeteria" cooking has sometimes been the best food I've eaten at the time. It very much depends on the circumstances and budget and your expectations. I've eaten like this in Leeds, London, Rome, France, Beijing. And not always in universities, the experience extends to factories, railways and even just the streetscape.
The FSU had a preference for ground meat and sausages at a state level because it was easier to ensure as much as possible of the meat inputs became food outputs. [This is wrong: they didn't] The Khrushchev flats had communal kitchens. They were universally hated I believe.
Ghost kitchens failed on lack of regulation and motivation to be the best they could be. There was never an intention to be "Jamie Oliver's ghost kitchen"
Re-intermediating consumers away from direct food carries on. Uber eats and Panda are delivering food from restaurants all around me, and one of them is a chinese take-away eponymously called "A Chinese Take Away" -I think it's a hack on the search engine carried forward. I've eaten their floating market soup in the shop, it's fine. Most of their trade is carry-out.
The main issue was that they relied on food delivery to sustain its model. I’m not going to pay $50 in food and $20 in tips and fees. Like anyone sane, I call in my order and pick it up myself. This idea died because of this missing link. I would entertain it if you have a pickup booth in a lobby.
And you just re-invented Chinese takeout as experienced in major cities!
One of my favorite dumpling shops in Brooklyn just had 2 tables... everyone just carried it out and back to home / work. And I think it worked well for them.
The emergence of ghost kitchens, more than anything else, is what got me to stop using food delivery services. They made it impossible for me to have enough trust, so I switched back to ordering from real restaurants that I physically go to.
I live a short bike ride from a CloudKitchen location and pick up from it often. Hard to beat birria and bao buns in the same stop even if they're nothing to write home about.
You end up standing in line with a bunch of delivery drivers who all know the drill and are on the clock, and you quickly learn you cannot be polite if you want to get your food.
Could be a few years stale. Some of the ghost kitchens even operated out of real restaurants by another name. E.g. higher quality sitdown place shoveling out burgers and fries out the back door.
When I worked for Dominos in the late 80s, it was a lot like this sounds. No dining room, though customers could walk in and order in a small vestibule. The place was as efficient as possible, just ringing phones, an assembly line, cooler and ovens, storage and cleaners in the back, and delivery drivers running in and out.
There seems to be something special about pizza that sets it apart from everything else, that made it seem reasonable to order it delivered from a non-restaurant even back then.
I think pizza is just virtually indestructible in terms of traveling. Sometimes I see americanized chinese food or those caribbean rotis being sold out of no-seating places likewise. The thing about a non-ghost no-seating establishment is you know they would go out of business eventually if they were truly awful. The ghost kitchens though can spin up new virtual brands endlessly.
It worked because Dominos was a brand name, people knew what to expect before ordering, and they picked up their own food instead of letting a overworked disinterested gig driver deliver it.
Most Domino’s were/are almost always delivery. Yes you can go pick up your pizza but most people don’t.
When I worked there they sold pizza and Coke. That was it. No breadsticks, no wings, no salads. One kind of crust, two size options. And by Coke I mean Coca Cola Classic in 16oz glass bottles. No Diet Coke, no sprite, nothing else. It was pure efficiency by elimination.
The drivers were all employees then, too. Not gig workers. No idea if that’s still the case.
Plenty of Chinese takeaways, and a good few “Indian” establishments (takeaway/delivery only, no restaurant) have operated in the same way, without chains or brand names, for decades, at least all over the UK. Many great quality, many poor, but that was part of the fun of moving to a new area, figuring out the good ones from all the menus that got shoved through the letterbox.
Before that, of course, the fish and chip shop is an ancient institution, though they rarely delivered.
The pizza was delivered by an actual employee of the pizza place, and there might be a small delivery fee and a tip. Now the gig companies add a delivery fee on top of the inflated menu prices, then ask for a tip before the order will even be picked up. The fees can be 80% or even higher than the in-store price.
I remember in the early 2000s there was a big push to deshittify delivery pizza. Companies were all advertising how they were now sending out their pizza in insulated bags. Dominos went particularly heavy, advertising a purpose built delivery vehicle with a built in warming oven (not sure if this was ever real or just for advertising) and a big emphasis on how they reformulated the entire menu to taste better.
When I worked there they had hot boxes drivers would carry in their car. There was an alcohol burner in it that somewhat kept it warm. They switched to insulated bags about a year after I started.
The stores usually had one or two company cars, a hatchback like a Ford Escort, painted up with the Domino’s logo. It was not equipped with any special pizza warmer. But most drivers used their own car. They got an hourly wage, a percentage of the order total if they used their own car, and tips.
Maybe Domino's had another delivery vehicle in the 2000s, but one of the 10 Tritan Domino's delivery vehicles from the '80s sold at auction a few weeks ago for US $45,000.
The vehicle was real. There is (or was) a series on YouTube about someone who bought one that had been salvaged and wanted to repair it for his own use but ended up getting various legal threats from Domino's, claiming he obtained the vehicle illegally or planned to misuse their branding.
For traditional restaurants, what percentage of orders are for delivery (using Doordash, etc)? Excluding pickups where the customer comes and picks it up himself.
I can't imagine it's even close to 50%.
I don't know the patterns of regular folks, but for me the prices in general have crept up enough that's it rare I want to try some new place - unless I get multiple strong personal recommendations for it. You can forget about paying extra for delivery to home!
> Quality control became impossible. Shared kitchen facilities meant that one staff member prepared food for multiple brands simultaneously. No ownership. No accountability. Just assembly-line cooking with zero connection to customers. When food arrived cold or wrong, customers had no relationship with the brand to forgive mistakes. No loyal regulars. No servers to smooth over problems. Just angry reviews that destroyed virtual brands forever. No reason for repeat business.
That's a straight quality control problem. It ought to be solveable.
But that's hard to do.
The trouble is, the whole food app industry is based on someone else dealing with the hard problems. The drivers aren't employees, and the restaurants aren't employees. If an app company gets into ghost kitchens, they are now in a business where they are clueless. Some try to avoid being responsible for the food by just being landlords for people who buy a station in the kitchen
Bulk food prep is a solved problem. Every major hotel has it solved. There will be some senior people who went to a serious culinary academy. They look at food prep as a manufacturing problem, with batch quantity optimization, holding time limits, error tolerances on temperatures and quantities, and quality control points. It's factory planning.
The first "Doordash Kitchen", in Silicon Valley, is near me. It's still operating, but I don't see many drivers there.
> When food travels twenty minutes in a bag, quality suffers.
You'd think that would be a solved problem in packaging and prep by now. Insulated containers are not rocket science.
You often want the opposite of insulation. Food continues to cook in the container, things get soggy, etc. Each dish and even ingredient can have different ideal packaging requirements. It's not something that really scales well. It's part of why menus like McDonalds' remained stable and relatively small over a period of decades. Notice how their fries are served in a specialty designed container that's open, which avoids them becoming soggy.
Isn't the main problem that the kitchens are shared and the restaurants are just popup brands with no loyalty built up?
I don't think you need an actual sit-down location to succeed, but you do need full control over what you produce and to spend time building trust and loyalty.
CloudKitchen was holding plenty of software hiring events near me few years ago. All these companies developed software to streamline the process of listing thousands of stores on all delivery apps, receiving orders, organize and manage kitchens assembly lines like how the orders are received, and dispatched to cooks, etc. Also they integrate analytics, cost tracking, supply chain management, and other random things like that. Basically for any PE or a billionaire who wants to larp as the next McDonalds or Starbucks, they don’t want to build everything from scratch.
The kinda thing a regular restaurant is probably managing using a spreadsheet and a notebook.
the basis of ghost kitchens started with one restaurant being cloned by the same restaurant to serve as a preparation nexus for their own menu to get local coverage and gain economies as the restaurant ramped up. Later pirate restaurants that were clones that copied the menu and snapped up the order - often assisted by delivery companies that had the stats. Later these grew larger, multi menu clones, often with 3-6 restaurant clones in one spot.
Then someone got greedy, and the rents for these spaces destroyed their economic basis, and the city got greedy, rubbing their oily hands together to get taxes, levy code shit etc. The original restaurants lawyered up and the ghost kitchens became anonymous - and Oh yes, COVID was over
they feel like scams and when I've accidentally ordered from a ghost kitchen it was by design a terrible experience.
I'm talking like, you order a 15$ main that is called "creamy pasta with prosicutto" and when it shows up its buttered spaghetti with a couple stamp-sized bits of ham. Ordering from actual restaurants come with some of the downsides the article assigns to ghost kitchens, like cold food and weird presentation, but ghost kitchens never seemed to reach the bar of "food someone would actually order, even if it was teleported to them instantly".
>>"food someone would actually order, even if it was teleported to them instantly".
The article states >>Quality control became impossible. Shared kitchen facilities meant that one staff member prepared food for multiple brands simultaneously. No ownership. No accountability. Just assembly-line cooking with zero connection to customers.
I'm not sure if it was impossible or if management never actually prioritized it, not bothering to understand what an actual customer would want. How much of it is the stupid management assumption that they can "just make a dish generally meeting description X on the menu" and deliver that and it'll be ok? «— Real question, did mgt fail at the product specification level, or was QC just as a practical matter, impossible?
On the economics, it really seems 30% for delivery is insane. It seems that same 30% might exceed the cost of the physical restaurant. And when it adds a 15-45min delay while homogenizing and cooling the meal, it seems an impossible problem. Maybe if the 30% transported it instantly and losslessly...
Probably good this soulless idea will die. Too bad so much perfectly good capital was squandered on it instead of better ideas
I think tech founders often underestimate what it takes to build a food business and what the margins are like and then start to cut corners to make the business viable.
And this is how it works in many (not all) American airports. Local restaurants put their brands on the signs, but the food is prepared by probationary employees of Acme Baggage Displacement And Cafeteria Management Corp.
But the airport newsstands that are just someone selling candy in a room with the name of a random local newspaper are an interesting local sight.
If an article makes it to the front page despite being AI-generated it probably has some interesting points or ideas, but it's unfortunate that people seeem to choose the speed and style of LLM writing over the individual style and choice that made the writing of yesteryear more interesting and unique.
Now I wonder whether any of it is even true.
You don't speak for us. If you are going to demand supporting evidence for obvious statements, then you can present supporting evidence for your spurious claims about value.
> No ownership. No accountability. Just assembly-line cooking with zero connection to customers.
> No loyal regulars. No servers to smooth over problems. Just angry reviews that destroyed virtual brands forever.
Pretty common pattern these days.
That, plus the hashtags at the end (unless Substack uses those and I was unaware of it), plus the fact that we know he's using AI in some capacity because of the feature images - it's a reasonable conclusion to draw.
I have never placed an order from a restaurant that I don't actually know beforehand. Maybe I haven't been there in person, but I know the name of the restaurant and I remember where is it and maybe someone I know ate there.
The only time I go completely blind is if I'm traveling. But that's like 1-2% of my total orders. And, in that case I usually stick to known chains or maybe I ask what's a good restaurant around here, then find it in the app.
I suspect it's not so uncommon to order from restaurants one already knows, rather than taking risks?
If that's the case, then ghost kitchens are going to have a very hard time getting their slice of the market
Eating out in US every meal would net $40-$60 a day. Multiplying by 30 days yields an ungodly amount of money. If you doordash it, then just multiply it by 2-3x.
I don't understand why there is no good cheap option. In Japan, I could go to a low-tier shop like Sukiya. It would cost $2-3; the food is decent with taste, and it fills my stomach. In US, $2-3 would be the min amount of tips.
Food that travels well requires different recipes, different ingredients, different packaging.
This doesn't ring true, given the popularity of DoorDash and the fact that most of those restaurant menus are in no way optimized for delivery.
Having lower-quality multitasking staff serving multiple restaurants from the same prep and equipment seems like an obvious way to lose quality. But simply streamlining the process of getting a menu onto delivery? That seems like a solved problem.
People have really funny ideas about restaurants. Somebody once left an online review of my family's establishment complaining that the hot chicken that was supposed to be on their cold to-go salad was in a separate container. They asserted that it was a "trick" to keep the chicken warm and moist, as though it would have been better to let the hot poultry heat their salad in the same container until it was lukewarm meat on top of wilted greens. Every day I wake up and mourn the IQ point that I lost reading it.
"Wilted soggy salad lovers hate this one simple trick!"
So some places are optimizing their fries for delivery.
I've also noticed some restaurants are better at adapting the packaging, like punching out ventilation so fried products don't steam themselves in transit. Lawrence Seafood (which rules) did that for a side of tempura we got this weekend.
But I agree in large part. I wouldn't order fried chicken delivered via door dash in any event. People doing that are optimizing for something other than quality.
This was an example of a well functioning ghost kitchen. I don’t know how profitable they were, but it was very convenient. There are a lot of downsides to this approach, like pre ordering, reheating, and limited menu, but it was a very different approach to current ghost kitchens around me now or DoorDash from a local restaurant.
I can't remember when I last used a service like that. The convenience isn't worth the disappointment and aggravation.
Are you sure they're not?
People are dumb.... but only for so long. I've been burnt so many times by delivery I've gone back to mostly ordering pizza and picking it up in person.
Oddly, "cafeteria" cooking has sometimes been the best food I've eaten at the time. It very much depends on the circumstances and budget and your expectations. I've eaten like this in Leeds, London, Rome, France, Beijing. And not always in universities, the experience extends to factories, railways and even just the streetscape.
The FSU had a preference for ground meat and sausages at a state level because it was easier to ensure as much as possible of the meat inputs became food outputs. [This is wrong: they didn't] The Khrushchev flats had communal kitchens. They were universally hated I believe.
Ghost kitchens failed on lack of regulation and motivation to be the best they could be. There was never an intention to be "Jamie Oliver's ghost kitchen"
Re-intermediating consumers away from direct food carries on. Uber eats and Panda are delivering food from restaurants all around me, and one of them is a chinese take-away eponymously called "A Chinese Take Away" -I think it's a hack on the search engine carried forward. I've eaten their floating market soup in the shop, it's fine. Most of their trade is carry-out.
Khrushchyovkas did not have communal kitchens; I grew up in one [0]. Perhaps you are thinking about kommunalkas [1]?
[0] https://news.ycombinator.com/item?id=7935844
[1] https://en.wikipedia.org/wiki/Communal_apartment
One of my favorite dumpling shops in Brooklyn just had 2 tables... everyone just carried it out and back to home / work. And I think it worked well for them.
Copy the address into your maps app and look it up on street view.
I live a short bike ride from a CloudKitchen location and pick up from it often. Hard to beat birria and bao buns in the same stop even if they're nothing to write home about.
There seems to be something special about pizza that sets it apart from everything else, that made it seem reasonable to order it delivered from a non-restaurant even back then.
When I worked there they sold pizza and Coke. That was it. No breadsticks, no wings, no salads. One kind of crust, two size options. And by Coke I mean Coca Cola Classic in 16oz glass bottles. No Diet Coke, no sprite, nothing else. It was pure efficiency by elimination.
The drivers were all employees then, too. Not gig workers. No idea if that’s still the case.
Before that, of course, the fish and chip shop is an ancient institution, though they rarely delivered.
The stores usually had one or two company cars, a hatchback like a Ford Escort, painted up with the Domino’s logo. It was not equipped with any special pizza warmer. But most drivers used their own car. They got an hourly wage, a percentage of the order total if they used their own car, and tips.
https://bringatrailer.com/listing/1984-tritan-a2-3/
I can't imagine it's even close to 50%.
I don't know the patterns of regular folks, but for me the prices in general have crept up enough that's it rare I want to try some new place - unless I get multiple strong personal recommendations for it. You can forget about paying extra for delivery to home!
That's a straight quality control problem. It ought to be solveable.
But that's hard to do.
The trouble is, the whole food app industry is based on someone else dealing with the hard problems. The drivers aren't employees, and the restaurants aren't employees. If an app company gets into ghost kitchens, they are now in a business where they are clueless. Some try to avoid being responsible for the food by just being landlords for people who buy a station in the kitchen
Bulk food prep is a solved problem. Every major hotel has it solved. There will be some senior people who went to a serious culinary academy. They look at food prep as a manufacturing problem, with batch quantity optimization, holding time limits, error tolerances on temperatures and quantities, and quality control points. It's factory planning.
The first "Doordash Kitchen", in Silicon Valley, is near me. It's still operating, but I don't see many drivers there.
> When food travels twenty minutes in a bag, quality suffers.
You'd think that would be a solved problem in packaging and prep by now. Insulated containers are not rocket science.
You often want the opposite of insulation. Food continues to cook in the container, things get soggy, etc. Each dish and even ingredient can have different ideal packaging requirements. It's not something that really scales well. It's part of why menus like McDonalds' remained stable and relatively small over a period of decades. Notice how their fries are served in a specialty designed container that's open, which avoids them becoming soggy.
Does anybody in the "ghost kitchen" industry?
I don't think you need an actual sit-down location to succeed, but you do need full control over what you produce and to spend time building trust and loyalty.
What does that even mean? Sell an NFT image of a burger?
The kinda thing a regular restaurant is probably managing using a spreadsheet and a notebook.