Hold The Storefront: How Delivery-Only “Ghost” Restaurants Are Changing Takeout

“Ghost restaurants” with no physical location are popping up across the country. Can a restaurant succeed if it’s delivery-only?

Hold The Storefront: How Delivery-Only “Ghost” Restaurants Are Changing Takeout
[Photo: Kate Zimmerman, courtesy of Green Summit] [Photo: Kate Zimmerman, courtesy of Green Summit]

Hungry New Yorkers ordering meals through such online services as Seamless or Eat24 order everything from sushi to burgers to tacos. But when they order from certain restaurants like Leafage and Butcher Block, they might not realize that those restaurants aren’t restaurants at all. They are virtual eateries created by a company called the Green Summit Group that operates several food-delivery services out of central commissaries in midtown Manhattan, Brooklyn, and Chicago. In New York alone, Green Summit’s brands offer all sorts of cuisine “concepts,” including meatballs, salad/sandwich/juice, and burgers/grilled cheese.


Green Summit is just one example of a growing wave of ghost restaurants that skip the storefront and bring food straight to the customer. These places range from the David Chang-backed delivery startup Maple to ready-to-eat meal services like Munchery to new business models like the integrated kitchen/delivery firm Good Uncle.

The Rent Issue

For Green Summit, which has an exclusive agreement with Seamless/Grubhub for delivery, there’s one major benefit to operating a virtual restaurant: You can’t beat the cheap rent. The company’s midtown Manhattan commissary at 146 East 44th Street has had a big advantage from the start: It doesn’t have to devote square footage to customer seating and waiting areas.

Green Summit cofounder Peter Schatzberg.

In a phone interview, Peter Schatzberg, Green Summit’s cofounder, says that a restaurant like Chipotle or Pret A Manger has to dedicate 75% of their space to seating, while 90% of their customers just grab and go. By comparison, a company such as Schatzberg’s can open inside a kitchen with as little as 200 square feet of space and operate a viable restaurant business with a minimal footprint.

Schatzberg and his partner Todd Millman started Green Summit in 2013 with approximately $1 million of investment. Their first online-only concept, Authentic (which still operates, see sidebar), launched in August of that year. According to Schatzberg, the restaurant made $20,000 in its first week of sales.

“To the consumer, it’s very much a kitchen,” Schatzberg says. “There’s an area with a grill, people working and portioning, and a room adjacent to the kitchen where orders are assembled. They’re all made to order, and stations are set up by category. For instance, there’s a sandwich area producing sandwiches for multiple brands. All the ingredients across brands are in the same areas, but you get specialization in staff where they focus on making salads and sandwiches, for instance. That’s all they do. It makes a better quality product, which ties into economies of scale.”


This business model works thanks to Green Summit’s reliance on external food-ordering platforms. In New York, a person ordering from Green Summit’s restaurants has to order through Seamless/GrubHub (which merged in 2013). The now joint company plays a crucial role in Green Summit’s growth: GrubHub loaned Green Summit money for their expansion beyond midtown Manhattan. According to Green Summit, it pays GrubHub/Seamless the same commission rates as other restaurants.

While, hypothetically, it would be possible for Green Summit brand to set up its own e-commerce websites and avoid the third-party fees, delivery services such as Seamless, Eat24, UberEats, and Doordash effectively serve as gatekeepers for any urban restaurant hoping to build a delivery business in 2017. In big cities like New York, Chicago, and Washington, most diners are more likely to open a delivery app on their phone than rummage through drawers for dog-eared delivery menus.

Not having an actual storefront means Green Summit can switch menus rapidly. Schatzberg mentioned, for instance, that the company quickly dropped a Middle Eastern concept for midtown Manhattan after encountering lower-than-expected sales.

It also means more versatility in the food they can offer. Similar or identical menu items show up across many of Green Summit’s online storefronts, and largely hews to a template of delivery-friendly dishes like salads, sandwiches, and grain bowls.

Offerings from Green Summit.[Photo: Kate Zimmerman, courtesy of Green Summit]

“There’s a lot of cross-utilization because at some point the universe of ingredients becomes finite,” Schatzberg says. For instance, when poke, a Hawaiian fish salad, became popular, he says Green Summit could jump on the trend because most of the ingredients were already in house for an existing sushi concept. “You don’t have to source new ingredients; you just train staff to cut fish differently or prep rice differently,” he says. “That ability to maneuver and build new brands is exponentially easier on a cost basis.”


Schatzberg estimates that it would cost traditional restaurants about $800,000 to test out new concepts like poke. Meanwhile, Green Summit told Crain’s New York that it loses as little as $25,000 if a new menu flops.

While Green Summit declines to offer historical revenue information, the company confirms plans to expand to New York’s financial district in February of 2017 and open a second Chicago facility in Q2 2017. Schatzberg says Green Summit aims to end 2017 with seven to eight locations and a run rate of approximately $25 million if same-store growth and location openings skew close to both projections and historical performance.

Good Uncle sets up agreements with established restaurants in other cities to replicate their menu items, like Joe’s Pizza.[Photo: courtesy of Good Uncle]

Marrying Restaurants With Logistics

Other companies in the ghost restaurant world are trying slightly different approaches. Maple, for instance (which Fast Company recently visited), and Sprig, have their own in-house delivery teams. But another startup is taking meal deliveries from commissaries one step further—and turning logistics into the center of the business model.

Good Uncle is a New York-based startup which is currently test-marketing at Syracuse University in upstate New York. The company sets up agreements with established restaurants with limited or no delivery service in Syracuse to license their recipes and then recreate them in the Good Uncle commissary. Users—college students are a current target market—then order meals through the Good Uncle app or through GrubHuband pick them up from one of several stops along a predetermined campus delivery route.

At press time, Good Uncle has licensing agreements with New York restaurants Croxley’s Ale House, Ess-A-Bagel, Joe’s Pizza, Sticky’s Chicken Fingers, and No. 7 Subs.

The Good Uncle app.[Photo: courtesy of Good Uncle]

The company recently raised $2.2 million; founder Wiley Cerilli is an early stage Seamless employee who later founded SinglePlatform before selling it for $100 million in 2012. Cerilli says Good Uncle’s cooks use the exact same ingredients as the restaurants they license menu items from and train with those outlets’ own cooks so dishes can be replicated as precisely as possible.

Good Uncle founder Wiley Cerilli

Good Uncle sells these meals at college student-friendly price points that range from $7 to $16 per item. And because customers pick up their food at a central pickup point, they pay no delivery fee.

According to Cerilli, the company keeps costs low through bulk delivery. In Syracuse, Good Uncle’s drop-off points are mainly in front of dormitories. “If it’s 6 p.m., they pick what they want to order, and the next page has a map showing the next delivery at the dorm will be at 6:45.” He points out the efficiency of this system. “Instead of in New York, where you deliver to my 40-story apartment building and then go to my apartment, and leave the building after—that takes a lot of time.” Good Uncle’s model for their delivery-only restaurant, he says, is “almost a bus route for the food.”

The kitchen in one of Green Summit’s commissaries.[Photo: Kate Zimmerman, courtesy of Green Summit]

A Future In Profitability?

Unlike India’s Dabba wallah system, which dates back to the 1890s, the concept of meal delivery is relatively new in the United States, only becoming widespread in the 1950s. Now, as the newest business models in this category, ghost restaurants are all grappling with a common quandary: How to turn a profit when the restaurant business is notoriously unprofitable and delivery is an inherently expensive operation?

A Cornell University study estimates that approximately 26% of restaurants fail in their first year of operations alone. While online delivery-only services might have lower overheads than a conventional restaurant, they miss out on the significant cash infusion that alcoholic beverage purchases provide… as well as the loss of walk-in foot traffic.


Leaked financials from the end of 2016 show just how tough a business this is. The documents revealed that Maple lost money on average on each meal in 2015, with the result of an operating loss of $9 million on $2.7 million in gross revenue. For 2016, the documents forecast operating loss of $16 million on top of $40 million revenue. Munchery weathered a difficult 2016 as well.

Of course, no restaurant will be in business long if it fails to deliver the single most important ingredient of all: the food itself. Good Uncle has the constant challenge of recreating the exact flavor of existing dishes far away from the restaurants they originated in. Green Summit’s always-try-new-concepts approach has had its share of hits and misses. Williamsburg’s Leafage, for instance, has four out of five stars on Grubhub, while Braised’s midtown location averages only two, with 33% of customers complaining of receiving inaccurate orders.

Opening up your delivery bag to discover someone else’s dinner? Maybe the solution to that all-too-common problem will come in the next phase of food-delivery innovation.