who ya gonna call?

Ghost Kitchens Will Always Be Dumb

They are not, in fact, the future of restaurants.

Who’s cooking? Photo: Education Images/Universal Images Group via Getty Images
Who’s cooking? Photo: Education Images/Universal Images Group via Getty Images
Who’s cooking? Photo: Education Images/Universal Images Group via Getty Images

You have probably heard that it is a bad time to run a restaurant. You may have read that 90 percent of New York restaurant owners weren’t able to pay full rent in December, or that many owners lose money when they can only fill their dining rooms to 25 percent capacity. You have seen the torrent of closings, and you know that many other businesses won’t make it till the end. The trouble is that it is very hard to profitably run a dining room for a base of customers who either can’t or won’t physically go in. Luckily, the tech world has come through with a solution for the ongoing hassle of physical reality: ghost kitchens. They are “the future of restaurants.” They “go the distance with delivery.” They offer a “brave new world of data-driven, search-optimized virtual restaurants.” A ghost kitchen is “a uniquely 21st-century innovation that promises to optimize & expand delivery service at (seemingly) minimal cost.” Ghost kitchens could “create a $1 trillion opportunity” in the next decade. They will catapult Guy Fieri from municipal government to “President of Flavornation.”

Ghost kitchens are “popping up everywhere” and yet they appear to be nowhere at all, which is exactly the point. The great value proposition of ghost kitchens is the illusion that they do not exist. They are designed to help facilitate the great magic trick of the last decade, which is that you can tap a button on your phone and 15 minutes later, someone shows up at your apartment with chicken wings.

Ghost kitchens are constantly billed as the next great restaurant innovation, an idea that first took hold well before the pandemic, and then really dug its claws in once the industry fell apart and potential customers grew accustomed to going days without leaving the house.

Ghost kitchens have been meticulously engineered to be infinitely adaptable and fantastically efficient. The Wall Street Journal loves them. But what they really are is a trend that manages, triumphantly, to strip away all joy from the act of eating. They are devoid of every feature that makes restaurants great, and they are not, despite what the many, many headlines say, the true future of the restaurant industry.

What is a ghost kitchen? Like ghosts — in this way, they are well named —the ghost kitchen can take many forms, but the overarching premise is that the ghost kitchen produces food to be delivered elsewhere, usually via app. Because there is no dining room, a ghost kitchen requires less staff and less space than a restaurant, and it can pivot on a dime. If the people do not want hamburgers, the ghost kitchen can rebrand as a taqueria. Perhaps it will do both. One San Francisco operator featured in a pre-pandemic rise-of-ghost-kitchens New York Times piece, runs three disembodied UberEats-specific concepts — fried chicken, burgers and wings, and ice cream — from the kitchen of his physical restaurant, which sells roast beef. It was going well, apparently. And good! Who can begrudge a restaurant for scraping by, especially in times like these?

In theory, the articles will tell you, ghost kitchens are a way for restaurateurs to experiment with new concepts. The lower startup costs could foster experimentation, evangelists say; this should be an exciting time. So far, though, in practice, ghost kitchens are mostly serving uncomplicated comfort foods, largely inspired by what tech platforms say people are searching for on apps. For now, “it’s a lot of chicken wings, a lot of grain bowls, and sandwiches and pastas, things that travel well,” allowed one restaurant strategist to Eater, insisting that the space is still “ripe for innovation.”

What ghost kitchens are not, as should be clear by now, are restaurants. They are food-logistics operations, and you could argue that is fine. Who cares if your fried chicken is from an abstract concept that is also an ice-cream parlor, which also sells roast beef? Are we to understand that fried chicken exudes an aura of pure chickenness only if it is prepared in a customer-facing space dedicated exclusively to poultry?

Obviously, no. But reading about the production of ghost sandwiches does not make me excited for the future of delivery. It makes me ache for restaurants. To eat a meal in a restaurant — even a bad meal, even at a fast-casual bowl chain — is to participate in an immersive experience. Things are happening. You are talking to another person, or you are eavesdropping on other people talking. You may be seated, for example, next to somebody who is extensively recounting their transformative experience with energy healing. You might chat with your server about their favorite foods. You might then take that server’s recommendation and find you have unexpectedly ordered a giant plate of curried bamboo shoots.

Delivery is many things: convenient, luxurious, sometimes festive, often disappointing, and at certain points during this pandemic, it’s all we have had. But delivery is neither spontaneous nor surprising. A delivery-only future is not the future that I want and, looking at the evidence, it isn’t a future that many other people want, either — since, historically, delivery-only concepts have perpetually failed.

In 2013, Green Summit Group launched a ghost kitchen in New York, burned through “hundreds of thousands of dollars a month,” and then shuttered in 2017. Maple, the David Chang–backed restaurant-food-without-a-restaurant startup failed, West Coast–based Sprig and SpoonRocket failed, as did Munchery. Chang’s own delivery service, Ando, which was supposed to reinvent the sad desk lunch with Momofuku magic, lasted two years before it was acquired by UberEats and buried forever.

Postmortems of these falls generally cite logistics — the slim margins of food service, the difficulty of delivery, the gulf between the realities of the restaurant industry and the expectations of big tech, ill-advised expansion — but there is another problem, which is less-often discussed: Delivery is not good. Delivery is, at best, okay. It is fine. It can even be enjoyable (especially if it is pizza). But compared to food served inside a restaurant, it is a constant letdown. It is gummier, soggier, oilier, messier, uglier, the wrong texture, the wrong temperature. When Eater’s Ryan Sutton tried Ando, he “ended up with a pile of very lukewarm, very un-crispy, very unremarkable strips of chicken breast.”

The greatest failure of Ando was that it could not transcend what it was: takeout. When you get takeout from a restaurant, though — even when it is disappointing, even when it pales in comparison to what it could have been — it at least gestures toward the real thing. It alludes to a location, a context, a history, and the people who have made this food. Just by existing, a restaurant makes its food better, or at the very least, more interesting. Without the restaurant, the food becomes the soggy equivalent of “content” — a McGuffin on which some business gets built, with no real concern for quality. The thing about food, though, is that if it is not good, people will not continue eating it.

Takeout, as Soleil Ho recently observed in the San Francisco Chronicle, is, in part, an argument for actual, in-person restaurants. A restaurant is a small adventure. My food memory is a catalogue of restaurant experiences: this olive oil cake, those pickled grapes, that Saudi prince sighting, those miserably uncomfortable chairs. But for the last year, food I haven’t cooked myself has come from a container, and been eaten at my house. I remember none of it.

Ghost Kitchens Will Always Be Dumb