These days, the weekday lunch rush at Singapore’s Chinatown Complex food center is more of a trickle.
Pre-pandemic, it was packed with tourists and office workers from the nearby business district. Today, it looks like a crime scene. Plastic webbing has been stretched over the tables to stop patrons congregating, and government-appointed “safe distancing ambassadors” patrol the floors.
More than a third of the 260 stalls are shuttered. At Jin Ji Teochew Braised Duck & Kway Chap, rows of poultry glisten under a heating lamp, the window surrounded by laminated copies of effusive reviews from local newspapers. It’s one of the few stalls that still has a queue outside: a couple of regulars and a solitary GrabFood rider waiting to pick up a delivery.
“We’re surviving,” owner Melvin Chew said. “Just surviving.”
In Singapore, hawker centers — where stallholders sell traditional dishes for a few dollars — are a rare egalitarian space in a city of high-rises and high rents, and are the city-state’s only entry on UNESCO’s Intangible Cultural Heritage list. Since the pandemic began, Singaporeans have rallied around the stalls through online “buy local” campaigns, and the government has supported them with subsidies and grants.
But Chew, who runs the 41-year-old business with his mother, Lim Bee Hong, said they are still in danger of dying out — wounded by the pandemic, but finished off by the food delivery platforms they shifted onto during successive lockdowns. Hawkers, he said, simply can’t compete on the apps, which typically charge 30% commission, wiping out any profit they could make.
“After you take out 30%, the hawkers can’t really profit,” Chew said. Most of them are firmly cash-based, and do their accounts on paper ledgers at the end of the month. “Only when they go through the procedure, they realize, ‘Oh, I earned nothing out of this.’”
Chew’s experience is typical. Rest of World spoke to more than a dozen owners of food outlets in Singapore, from hawker stalls to high-end restaurants. They all felt trapped: While the platforms allowed many to remain open and reach their customers during lockdowns and curbs, they did so on terms that restaurants say are unsustainable and risk cannibalizing their businesses. As restaurants close down, some are even being subsumed into the delivery platforms themselves, their kitchens repurposed for “virtual brands” — hinting at a future where restaurants compete with the platforms’ own private-label products.
The main complaint among restaurateurs, however, is that this disruption is not the result of fair competition or successful innovation. It is, instead, a function of the massive amounts of capital being deployed in search of a viable business model.
“[The platforms] are willing to sustain these losses for years and years and years. But in the end, it’s the restaurants that are suffering,” said Sid Kim, who runs three restaurants in Singapore. “We can’t pay these exorbitant amounts for years and years and years. We can’t lose money on every delivery order for years and years and years.”
Kim’s flagship restaurant is an upmarket Korean-Mexican cantina, Vatos Urban Tacos on Beach Road. It was one of the first five outlets on Uber Eats when the platform launched in Singapore in 2016.
Kim was offered a discounted commission rate of 25%, and the platform threw money at promotion. Business was brisk. Soon, he said, he was bringing in $30,000 to $37,000 (40,000 to 50,000 Singapore dollars) a month on delivery.
Two years later, though, when Uber sold its Southeast Asia business to Grab and UberEats was replaced by GrabFood, Kim’s earnings dropped precipitously.
In order to optimize riders’ routes, the platforms reduced his delivery radius, cutting him off from wealthier, more international districts around Orchard Road, River Valley, and Tanjong Pagar.
“They just said, ‘This is your zone,’” Kim said. Under the new system, he said, his restaurants saw 10% of the orders they were accustomed to on Uber Eats.
He stuck with the platforms for a while but began to conclude it wasn’t worth the time and effort. “And then Covid happened,” he said.
Since April 2020, Singapore has imposed limitations on indoor dining, from total closures to restrictions on how many people can eat together. Restaurants scrambled to get online. Some jury-rigged their own delivery services, hiring out-of-work taxi drivers or booking third-party logistics providers like Lalamove or Pandago.
But most went where the customers were — Deliveroo, Foodpanda, and GrabFood, which have been joined in the market by smaller competitors, including Oddle, Bungkus, and Chope. Some restaurants told Rest of World they were paying as much as 40% in commissions.
The platforms didn’t just have scale. They also had money to burn, allowing them to subsidize delivery fees by up to half.
“They’ve created this unrealistic expectation,” Howard Lo, founder of Empire Eats, which operates six bars and restaurants in Singapore, told Rest of World. “I think the subsidized pricing … has distorted the market.”
Venture-backed gig work platforms typically use subsidies to build market share. In the food business, they’re pushing the cost of discounts onto the restaurants.
Not long after Singapore went into its quasi-lockdown in early May, Kim got a call from a sales representative at Foodpanda, who told him that the company had noticed he wasn’t making much revenue through their platform.
It wanted to offer him a deal: If he agreed to run a 20%-off promotion, his restaurant would be displayed on the main page of their app. With thousands of restaurants crowding onto the platform, it can be the only way to stand out or to access special perks like islandwide delivery.
“I was like: Wait. You want me to pay the 30% commission and the 20% promotion?” Kim said. “‘So you want me to pay 50%? Why the hell would I want to do that?”
Paying 50% off the top would guarantee he’d lose money on every order. “And she goes: ‘Ah, but it’s to increase your sales,’” Kim said, incredulous. “Why would I want to increase my sales when I’m losing money? The more sales I have, the more money I lose.”
Other restaurant owners gave accounts of similar conversations, illustrating the disconnect between how VC-backed tech companies pursue success and what businesses need to do to survive.
“You want me to pay the 30% commission and the 20% promotion? So you want me to pay 50%? Why the hell would I want to do that?”
A Foodpanda spokesperson said that the company had rolled out several policies to support restaurants, including waiving commission fees for hawkers during the most recent monthlong lockdown, offering free photo shoots to vendors, funding deals for merchants, and free delivery for hawker center orders above $25.00 (35 Singapore dollars).
A spokesperson for Deliveroo outlined similar support, including free delivery for some outlets during the circuit breakers. They said that restaurants on the platform “on average, see their revenues increase by 30% or more.”
The companies did not disclose how many restaurants have taken advantage of these policies. None of the restaurant representatives who spoke to Rest of World for this piece said that they had.
The Deliveroo spokesperson defended its commission rates of 30%–40%, noting that they went toward “paying our riders fairly and keeping them safe on roads” as well as covering the “full spectrum” of services involved in running a food delivery platform, operating 24/7.
Both Lo and Kim, along with several of the other restaurants that Rest of World spoke to, have recently had new offers from the delivery platforms, which hint at a new strategy.
Most of the major apps have begun opening cloud kitchens and ghost kitchens — spaces where brands can make food specially for delivery, often located in residential areas closer to customers. Deliveroo has two cloud kitchens in Singapore, and GrabFood two.
Some have worked with entrepreneurs using their cloud kitchens to create virtual brands — “restaurants” that exist exclusively or mostly on their platforms. On the platforms, they look indistinguishable from any other outlet and usually offer generic midpoint versions of national cuisines — Italian, Mexican, Greek, or Chinese.
Cloud kitchen operators, including the delivery platforms, have started reaching out to struggling restaurants and suggesting that they turn over their underutilized kitchens to churn out meals in a kind of franchise model.
“They say, we know you have a Mexican restaurant, why don’t you do this Mexican brand … because you already have the ingredients and stuff. Why would I do that?” Kim said. “I spent 10 years building my Mexican brand. Why would I take the Mexican brand you built last week and stick it in my portfolio?”
These virtual brands have lower overheads than brick-and-mortar restaurants, and they have access to the platform’s data, allowing them to optimize their menus and locations. Restaurant owners told Rest of World their concerns that if the platforms are making higher margins on virtual brands, they have an incentive to promote them on their apps, making it even harder to compete.
This is a model that has been successfully, if controversially, used by Amazon, which promotes its private-label products at the top of its search rankings. It has been accused of replicating popular products and using its monopoly on data to gain an advantage.
A Grab spokesperson said that cloud kitchens allow “merchants to experiment with new ideas in a low-cost and low-risk manner” — for example, by testing out virtual brands to see what works for a local user base. “Cloud kitchens also help us to bridge cuisine gaps in specific areas where we see high unmet demand,” the spokesperson said. Grab, through its GrabMerchant platform, does share data with restaurants so they can see which items sell well, helping them to optimize their menus and marketing, the spokesperson added.
A Deliveroo spokesperson echoed the same, adding that the company sees virtual brands as ways for restaurants to reach new customers and test new menus, “but not with the intention of competing with their existing business.”
GrabFood isn’t profitable. The company is preparing a SPAC listing on Nasdaq later this year. Deliveroo went public in April in what the Financial Times described as “the worst IPO in London’s history” after its shares fell 25% on its first day of trading. In its pre-listing filing, the company posted a loss of $309 million. Foodpanda, which is owned by the German company Delivery Hero, reported more than 2.5 times growth in revenues in 2020 but still made a loss.
The desire to move further up the value chain is motivated by the desire to impose some efficiency on the “patchwork of systems” that brings offline businesses — restaurants and logistics — online, according to Howard Yu, the LEGO professor of management and innovation at IMD business school.
According to Yu, the platforms have two options. They can try to build a constellation of services where lower-margin businesses, like food delivery, generate data and drive customers toward more profitable ones, like financial services. The alternative is to use their data to build vertically integrated models, like Domino’s Pizza, the US pizza chain, whose ownership spans the brand, sales platform, and logistics businesses but has franchisees running the actual kitchens.
Grab is firmly in the former category. It began in ride-hailing, expanded into food delivery, and is now investing heavily in financial services, including payments and more complex products. Restaurants on Grab’s Merchant platform can already apply for working capital loans through the system. Their creditworthiness is assessed using their takings on GrabFood. Deliveroo and Foodpanda have both moved into grocery delivery, allowing them to use their logistics networks for a new set of products.
Yu observed that there were too many platforms and that consolidation loomed. The platforms’ pricing is going to have to change, too. “Right now, it’s subsidised by venture funds,” he said. “At some point, consumers need to pay. And that’s a matter of fact.”
In the short term, the competition between platforms might seem good for customers, but it’s bad for the restaurants caught in the middle, and the damage it does could be permanent. Restaurants aren’t just businesses but part of the cultural fabric of cities. There have already been high-profile casualties among Singapore’s heritage restaurants.
Cedric Tang grew up in his family restaurant, Swee Kee Eating House, on Chin Chew Street in Chinatown. His grandfather opened his first restaurant in 1939, became a roadside hawker after World War II, and finally settled down again in permanent premises in the 1950s.
The restaurant, which specialized in Cantonese-style fish soup, was on the ground floor, and Tang’s family lived on the third. By the time Tang took over as managing director in 2018, the family had opened two other restaurants in Singapore and two others in Indonesia, and Swee Kee Eating House had relocated to Amoy Street, a trendy area of the central business district.
In May, Swee Kee Eating House closed down for good. “I was thinking to myself, without our oldest restaurant, who am I?” he told Rest of World.
Tang said he couldn’t find a way to make delivery profitable. Many of his staff had been with the restaurant for years and struggled to adapt to technology; some didn’t speak enough English to use the interfaces.
Thirty percent commission, the cost of promotions, the fact that some of their popular dishes wouldn’t travel well, the manpower needed to service the orders and manage the technology meant that his only option was to close. Since making the announcement, Tang has been overwhelmed with messages from his regular customers, who are bereft at the loss.
“This is the thing that I think a lot of people forget — the ultimate winner is going to be the … platform,” he said. “The ultimate loser is the consumers.”
