Rescuing Restaurants: How to Protect Restaurants, Workers, and Communities from Predatory Delivery App Corporations

September 18, 2020 Anti-Monopoly Policies & EnforcementCOVID-19 EconomyTech

At the end of March 2020, about two weeks into the national coronavirus lockdown, three of the leading food delivery apps were struck by the same marketing strategy, aligning themselves with a mission to “save” independent restaurants. Postmates shot an ad campaign titled “#OrderLocal” featuring celebrities like Mindy Kaling saying, “You don’t want to come out of this tough time and find that all your favorite small businesses are closed.” DoorDash launched a campaign called “Open for Delivery,” temporarily waiving the delivery fees charged to consumers, and later followed up with an ad campaign featuring celebrities like George Lopez and Ming-Na Wen talking about restaurant jobs they had before they were famous. Grubhub also rolled out a promotion called “Supper for Support,” exhorting its 23.9 million users to rally around the small enterprises that “are the lifeblood of our communities” and promising a $10 discount on any order placed between 5 p.m. and 9 p.m. as a kind of reward for their solidarity, “so you can save while supporting the restaurants you love.” 

Few outside the restaurant industry could appreciate the way these marketing campaigns misrepresented the economics of restaurants. The four dominant delivery apps—two of which have since announced they intend to merge with one another—charge large commissions to restaurants for the service of processing orders, even more for delivering them, and still more for lending promotional support like the $10 discounts. These fees virtually guarantee that all orders placed to independent restaurants over delivery apps are unprofitable for the restaurants. 

Before the pandemic, many chefs and small operators paid the fees, believing they lacked the leverage or power to do anything else. But once the pandemic shut down most dining rooms, wiping out more than $145 billion in restaurant sales between March 1 and June 30, thousands of chefs realized that staying open to feed the delivery apps would bankrupt them. 

In San Francisco, the local independent restaurant association began pleading with the apps for a 50% break on commissions, reasoning that surely 50% commissions from all the restaurants would be preferable to 100% commissions from no restaurants at all. But the corporations refused, instead hatching more predatory cash extraction schemes like “Supper for Support,” which required restaurants to shoulder the entire $10 discount, and also pay Grubhub its commissions on the full pre-discounted price of the transactions, so that a $32 order would yield just $12 to the restaurant. 

By mid-April, many city governments had begun to fight back, using emergency orders to cap the fees the apps were allowed to charge. Over the next three months, state and local legislators in Seattle, San Francisco, Los Angeles, New Jersey, New York City, Washington D.C., Philadelphia, Cincinnati, Portland, and many smaller municipalities passed laws capping the fees third party delivery apps were allowed to charge restaurants during the pandemic at maximums of between 10% and 20%. Many of the laws passed with unanimous support from local legislators. 

The apps’ response to the rash of new laws was illuminating: in city after city, they simply refused to acknowledge them. Postmates was the most consistent in its disregard for the law, flouting the first caps passed in Seattle then Los Angeles and Washington D.C., and finally Portland, where the smaller delivery app was joined in its noncompliance by its much larger rival Grubhub. 

In San Francisco, where DoorDash’s 64% market share is more than quadruple that of its next biggest rival, the company admitted to continuing to charge 30% commissions for nearly three months following Mayor London Breed’s April 10 institution of a 15% fee cap. DoorDash ultimately agreed to pay back the commissions, while Grubhub, after initially complying with California’s caps, increased commissions again in June, telling partner restaurants the fee cap had expired once local laws allowed them to open for patio dining.

Restaurants are increasingly struggling to fend off this deep pocketed clique of tech companies determined to use them purely as vehicles for extracting fees and consumer data, which together with an unprecedented shutdown of dining rooms now threaten their industry with extinction. And far from providing great service to consumers, these delivery apps have often failed miserably to improve on the old pizza delivery guys; one survey of regular app users found that cold food and order inaccuracies led them to report being dissatisfied with more than a quarter of all delivery app transactions. 

This paper will discuss the ways delivery apps have used abusive, deceptive and often illegal practices to take control of the restaurant industry, and how their business model increasingly poses a threat to the future of American restaurants. It will provide case studies of Grubhub and Doordash to explain the similar but distinct models these companies have used to build their dominance. And finally, it will provide solutions policymakers could pursue to ensure that independent restaurants aren’t all subsumed by these predators.

Behind the Delivery App “Wars”: A De Facto Cartel 

The existence of four significant players (two of which intend to merge) has led the business press to describe the restaurant delivery business as “fiercely competitive.” This competition, however, has not been oriented towards creating a better service, but towards capturing market power. And they are willing to lose money to do so. Not one of the third party delivery apps is currently profitable, and three never have been. DoorDash alone lost nearly half a billion dollars in 2019. 

What the apps have done, instead of competing to serve customers and restaurants, is use Wall Street money to accumulate market power, raise barriers to entry, and then merge with each other and set up regional monopolies.The people who have invested tens of billions of dollars in the four dominant delivery apps tolerate huge short-term losses purely because they see the likelihood of monopoly power.

Sometimes those backers are the same funds. Both Uber and Grubhub received substantial early-stage investments from both Benchmark Capital and its founder Bill Gurley, who served on the boards of both companies until Uber launched UberEats in 2015. Benchmark co-founder Andy Ratchleff was a seed investor in DoorDash. Softbank invested $7.6 billion in Uber and nearly a billion dollars in DoorDash, and last year, the Saudi-backed Japanese venture capital firm ordered DoorDash and UberEats to conduct merger talks. Pending its acquisition by the Dutch delivery app Just Eat Takeaway, Grubhub’s third-largest shareholder is large fund manager BlackRock, which also holds a significant minority stake in Postmates pending its own acquisition by Uber. 

But the most compelling evidence of the harmful type of competition involved in the delivery app “wars” is the effect this multibillion dollar influx of funds into restaurant delivery apps has had on the commissions restaurants pay for presence on the platforms and the wages couriers make for delivering the food. The former, which were arguably high when Grubhub was a dominant player in most markets in 2014, are substantially higher now; while the latter, in spite of a 20% growth in the food delivery sales over the past 5 years, fell by some estimates dramatically short of minimum wage.

Food delivery apps are match-making institutions whose appeal is based on their ability to provide customers to restaurants and restaurants to delivery drivers and customers. The key mechanisms to dominating a city are featuring a broad enough range of restaurants while maintaining a large enough pool of drivers that consumers know they can probably order what they want, when they want it using the app, which then, theoretically, encourages more restaurants to sign up because the app is what all the customers use. 

Instead of competing for restaurant partners to grow its marketplaces, DoorDash and Postmates used deceptive practices and industrial scale menu plagiarism to simulate the appearance of official partnerships, filling its marketplaces with the offerings of restaurants whose owners had no idea the apps even existed. When orders came through, call center employees hired by the apps would order the food manually and couriers would pick it up under strict orders to not identify themselves. When the restaurants caught on, invariably because the deception had resulted in angry phone calls from customers — most frequently because the menu listed on the platform was out of date and the dishes ordered were no longer available, or because the chef had prepared the food for trip up an elevator and not a 45 minute bike ride — the restaurant management would contact the offending app attempting to get their menu removed from or updated on the platform, and a salesperson would claim to be powerless to do anything about it unless they signed on to an official partnership. 

These practices are deceptive, predatory and in many cases reliant on intellectual property theft, as a Santa Fe restaurant owner detailed in a recent open letter to a delivery app CEO:

“If we don’t sign up for this ‘partnership’ you pirate our menus off our website and take orders from customers anyway. The pre-charged payment cards sometimes don’t work and everything we made languishes, unpaid for. We field angry calls from customers who think it’s our fault they didn’t get the food they ordered. When my manager called customer service to tell you how unfair it is that we are paying for your mistakes, he was told ‘Well, none of this would happen if you would just sign up with us.’ Which sounds a lot like what the mob boss says after they burn down your house.

The contracts the app salespeople induced restaurants to sign in most cases included anti-competitive “No Price Competition Clauses,” requiring restaurants to keep prices the same across all food delivery apps and in-person dining. These clauses had two consequences. First, they triggered a shift in customer behavior toward relying on delivery services in lieu of showing up in-person. Second, by ensuring price uniformity across all apps the clauses indirectly absolved the apps from competing with one another, either for users or restaurant partnerships, on the basis of price; if consumers could be sure prices would be the same regardless of which platform they ordered from, there was little incentive to offer restaurants lower commissions in the hope they would do more business over the platform, and at the same time price uniformity made it easier to sway customers with the use of special delivery discounts, rebates, loyalty clubs and other promotions—for which the delivery apps could then in turn charge restaurants additional fees. 

Similarly, instead of competing to hire drivers and bike couriers to deliver their orders, DoorDash, Postmates and Grubhub blanketed each market they entered with recruiters pitching the jobs as supplemental income sources, with the intention of misclassifying their workers as “independent contractors,” exploiting a loophole in labor law that allows corporations avoid the legal responsibilities of hiring employees simply by calling those workers contractors. Aided by referral bonus programs and promises of $2,000 weeks,, the apps flooded their systems with so-called “gig workers,” depressing wages, then gradually pared and professionalized its amateur workforce by assigning regular schedules, penalizing workers who refused jobs that took them too far or paid too little and using extreme micromanagement and surveillance technology to exert increasing control over their “independent” staffers.

But as their drivers began to resemble a traditional workforce, the delivery apps continued to misclassify them as contractors, invoking an oft-cited industry claim that 80% of app-based drivers “work only part-time” despite the Bureau of Labor Statistics estimates that roughly 73% of so-called “gig” workers do the job full-time. (The industry figures, a former lobbyist for multiple app companies told the American Economic Liberties Project, are misleading because they only count workers as “full time” if they work full time for a single app, despite surveys suggesting that more than 80% of app drivers work for multiple apps.

Misclassification enables the delivery apps to pay rates far below minimum wage, once mileage and additional payroll taxes are deducted. One analysis found that DoorDash drivers took home an average of $1.75 an hour after taxes and mileage were factored in; the same study found that drivers actually lost money on fully a third of jobs. Frustrating matters, the apps conceal much relevant information—the ultimate location of a given job they are asked to accept, for example—from drivers, chronically tweak their formulas for remunerating couriers, usurp tips to apply them to guaranteed minimums, and levy fees and adjust policies in ways that keep workers constantly struggling to stay afloat. DoorDash even uses its workers to extract additional fees, charging workers who elect to receive their earnings at the end of each shift $1.99 for the privilege, and more recently charging “Dashers” shipping fees as high as $43 for “free” hand sanitizer and gloves to protect themselves and customers during the pandemic.

The apps, especially DoorDash, have deliberately flouted laws passed to protect their workers.

Most notably, DoorDash has taken no apparent steps to comply with a California law that went into effect in January 2020 codifying a 2018 state supreme court decision against an Amazon contractor that requires companies classifying laborers as independent contractors to meet three strict thresholds to prove they were not employees. Nearly six thousand DoorDash couriers filed arbitration claims demanding to be reclassified in six weeks after it passed. 

Together with Postmates, DoorDash instead pledged to spend $90 million repealing the law with a ballot initiative called Proposition 22. In a June lawsuit against the company, the newly-elected San Francisco District Attorney Chesa Boudin detailed the numerous ways DoorDash micromanages the time of its couriers and described DoorDash’s defiance of the law as “calculated decision.” Certainly it was a costly one: over and above its share of that $90 million, the company risks a penalty between $5,000 and $25,000 for each infraction. 

Making matters more difficult for both workers and restaurants, the delivery apps have established regional monopolies that effectively suppress competition especially in the realm of wages. DoorDash maintains a 65% market share in San Francisco, diminishing the competition for courier labor in a pattern that repeats itself throughout most of the biggest metropolitan markets, half of which are more than 50% controlled by a single one of the four dominant players. 

As discussed above, the delivery apps are serial mergers. Grubhub and DoorDash alone comprise more than 20 companies that once competed with one another. Grubhub in particular is a classic “roll-up” of a few menu portals, a point-of-sale systems integrator and about a dozen regional food delivery services, most of which raised prices immediately upon being acquired. investors plowed nearly a billion dollars into Postmates despite its single-digit market share solely because of its potential “takeover interest.” Before Uber consummated that interest, the cash-burning rideshare giant also held merger talks with both Grubhub and DoorDash. “The constant refrain you hear from all of these companies when they go on CNBC or talk to investors is that they’re going to start making money as soon as the next merger closes,” says a former executive for one of the major delivery apps. 

Finally, three of the four dominant apps have also carved out an unfair advantage for themselves vis a vis any law-abiding competitors by cheating the government. Although GrubHub has paid hundreds of millions of dollars in sales taxes on its commissions over the course of its existence, neither DoorDash, Postmates nor UberEats have ever paid sales taxes on theirs, according to a 2019 investigation. Perversely, in a practice chronicled extensively in online forums, DoorDash did collect sales taxes on the food it sold in three states—Delaware, New Hampshire and Montana—that do not tax food that is consumed off-premise. DoorDash simply pocketed the ostensible tax as extra revenue. 

Although all the delivery apps have pursued a strategy of market domination, some important differences in their corporate governance can help contextualize their sometimes-divergent approaches to that dominance. Grubhub and DoorDash specifically are the products of markedly different formative financing cultures. And it is worth discussing them briefly because they happen to be the dominant sources of financing in American business, both involve their own unique antitrust implications, and in the case of Grubhub and DoorDash they have worked somewhat symbiotically. 

Grubhub: The Walled Garden of Restaurant Extortion

The best way to understand the perniciousness of Grubhub’s business model is to Google a restaurant, preferably a mid-tier one from which a lot of people might order takeout. Invariably, you will call up a long list of websites either owned by or in business with the delivery apps, and for most of them that delivery app is GrubHub, which owns Menupages, Allmenus and Seamless, along with Eat24, a former subsidiary of Yelp which remains Yelp’s “official delivery partner.” For years ,thousands of restaurants also found themselves competing against doppelganger websites owned and designed by Grubhub, which spent much of the last decade buying up tens of thousands of URLs related to real restaurants, outfitting them with rudimentary websites that directed users back to GrubHub, and using search engine optimization to move them sites up the search results. The result is that a casual restaurant’s actual website is sometimes not near the top of its own search rankings, which leaves those restaurants at the mercy of Grubhub.

Grubhub’s domination of the restaurant internet is both inspired by and inextricably linked to Google’s monopolization of search.

Founded in 2005 alongside the Huffington Post and Bleacher Report, Grubhub joined a wave of internet properties that learned to build audiences by mastering Google’s algorithms. Like Google, Grubhub originally billed itself as a search engine, but over the years increasingly blurred the lines between paid and “natural” content until it became impossible for an enterprise to avoid getting buried in the search results without paying up. (The FTC has repeatedly and explicitly ordered search engines to clearly differentiate paid from unpaid content, but its guidance has persistently failed to police the ever-subtler ways Google advertises to its users.) And as Google has metamorphosed into what House Antitrust Subcommittee Chairman David Cicilline termed a “walled garden” of corporate profit, Grubhub has used Google to appoint itself as a kind of restaurant toll collector. 

More often than not, for example, the “takeout” or “delivery” button displayed on the official Google Business listing of a small restaurant will lead the user directly to Grubhub, with each click resulting in a fee for Google just as each order results in a larger fee for the delivery app. It is not clear how much Grubhub pays Google for linking its official sites to the restaurant portal, but a bar owner who tried to get his Google Business listing to link to his own delivery service told Economic Liberties that he was initially quoted a price of a dollar per click regardless of whether the user ordered anything.

Grubhub goes to great lengths to extract tolls from its restaurants. Until last year, the company listed phantom restaurant phone numbers on its networks of fake Grubhub sites, which would route callers to the restaurant’s real number while enabling Grubhub to charge the restaurant a commission for the “lead.” A class action lawsuit detailing the practice revealed that Grubhub had charged one small chain thousands of dollars for phone calls that had never led to transactions; Grubhub was simply charging restaurants for every call lasting longer than 45 seconds. The lawsuit was thrown out for violating the arbitration clause of the contract in which Grubhub claimed it had obtained the restaurant’s permission to charge for the phone calls, but not before numerous other restaurants discovered thousands of dollars in dubious phone charges on their invoices.  

Over and over again, Grubhub has justified its practices by characterizing them as “marketing services” provided to restaurant partners with their explicit permission to “help” them navigate the internet. This seems unlikely; no user Googling a restaurant by name needs that restaurant to be “marketed” to them. To the contrary, what Grubhub actually achieves by proliferating all these Grubhub-owned restaurant internet properties is simply the ability to take a cut out of every transaction a restaurant completes over the internet, charging that restaurant over and over again to access its own customers. These marketing services regularly drive commissions as high as 65%. 

The company also uses its dominance in search to retaliate against restaurant owners who cancel, downgrade, or refuse to sign contracts with Grubhub. A Grubhub client restaurant in New York told that her order volume fell off whenever she attempted to opt out of the premium services that had led her Grubhub bill to exceed her rent. A Miami pizzeria that had cancelled its Grubhub contract and built its own online ordering platform told one journalist that Grubhub had adjusted all its websites to tell users, inaccurately, that their restaurant was not taking orders. Economic Liberties spoke to another restaurant owner in Portland, Oregon who said she was terrified to close her account precisely for this reason. Similar to the deliberate proliferation of fake restaurant websites and fake restaurant phone numbers, it may also violate federal laws barring unfair and deceptive practices.

“Grubhub is unique among the delivery apps in that it has literally no sense of optics,” one former company executive told Economic Liberties. It is also the only one of the apps that has ever turned a profit. Much of its corporate identity has been shaped by the private equity industry, which effectively acquired Grubhub in its 2013 merger with the private equity-owned Seamless. Under this management, Grubhub fostered what the executive calls an “almost comically aggressive corporate culture” along with a classic private equity operational strategy known as the “roll-up,” wherein a parent company acquires a string of companies in the same line of business in the theoretical pursuit of “economies of scale”—and in most cases, the actual pursuit of the opportunity to raise prices. After buying the OrderUp platform from its parent Groupon in 2017, Grubhub doubled the fees it charged restaurants to 30% from 15%, acccording to a group of Iowa City restaurant owners who responded by launching a rival service. 

Like the innumerable private equity-led rollups in the health care sector that consolidated over the same time period, Grubhub accompanied its acquisition spree with aggressive across-the-board price hikes, jacking up commissions by eight percentage points between 2014 and 2019. Asked whether he would consider yet another merger in the interest of expanding his profit margins on a CNBC appearance in February, Grubhub CEO Matt Maloney became defensive. “We’ve bought ten companies in the past ten years, I’m very opportunistic,” he said. “There are a lot of companies losing a lot of money and the industry is ripe for consolidation.” Three months later, he announced he would be selling the company to the European delivery app Just Eat Takeaway.

DoorDash: Industrial-Scale Trademark Infringement and Big Dining Data

The youngest of the dominant delivery apps, DoorDash quickly became the biggest of all the delivery app companies thanks to the easy Silicon Valley money and “opportunistic” approach to the intellectual property of others that has defined its existence. It was founded by Stanford Business School students at the famous YCombinator “startup accelerator” camp in Mountain View, California, in 2013, the year one venture capital coined the term “unicorn” to describe the increasingly common phenomenon of startups valued at more than a billion dollars. DoorDash achieved unicorn status within the first three years of its existence and ultimately raised $2.5 billion, nearly ten times the amount of funding Grubhub ever had. 

While DoorDash has borrowed tactics from many companies over the course of its short life, the core of its business model is cribbed from Amazon.

By the middle of the decade, Amazon had spent billions of dollars subsidizing free and fast delivery in a strategy that transformed consumer behavior, solidified its domination over all retail, and taught investors that long-term losses could yield dividends in the form of market power. DoorDash bet that it could pull off a similar strategy with restaurants by combining the logistical platform and workforce of a company like Uber with the revenue model of Grubhub. 

By 2019 DoorDash had eclipsed Grubhub’s share of the restaurant delivery market, thanks largely to a move the company borrowed from Postmates: offering consumers delivery services from restaurants that did not offer delivery and were unaware of the existence of this new Silicon Valley subsidized business model. DoorDash developers scraped the internet night and day for menu content the company then would quickly repurpose into unauthorized DoorDash “microsites” to give users the illusion of a full, comprehensive network of choices. Nor did it stop at scraping menu items and prices. According to lawsuits filed by the restaurant chains Legal Sea Foods and In-N-Out Burger, DoorDash had illustrated its menus on its website with slightly altered versions of their logos and in the case of In-N-Out, DoorDash had also sent employees to its physical locations to blanket the stores with fliers boasting about the new delivery service. Orders that came through the app’s platform were then manually placed by call center agents impersonating customers and delivered by couriers who were specifically instructed to leave their branded company gear in their cars. 

As the lawsuits explain, restaurants resented DoorDash for its unauthorized deliveries because the people who used the app were likely to post reviews on Yelp and TripAdvisor if their orders showed up cold or somehow incorrect—or not at all, because the platform’s scraping algorithms had failed to detect that the restaurant in question was actually closed. And while arguably less harmful than taking a third of their revenues, DoorDash exploited the widespread frustration with its “disruptive” services as an opportunity to sell restaurant owners on a formal partnership, wherein DoorDash would make some recommendations about increasing order volume and send its couriers with insulated bags in exchange for a 30% commission on all the sales it processed. 

In formalizing its relationship with restaurants, DoorDash shifts the burden of subsidizing the delivery from its deep-pocketed Silicon Valley backers to struggling restaurants, in much the way Amazon has shifted the tremendous costs of subsidizing one and two day delivery to Prime customers onto its third party vendors, who are now all but required to use the platform’s expensive fulfillment services if they want to maintain sales on its marketplace. 

But subsidizing cheap (and anonymous) delivery services is not in the long-term interest of restaurants, which survive on the strength of their personal relationships with consumers—relationships that the third party delivery apps sever, usurp, then all too often monetize for their own purposes. Which brings us to the existential threat DoorDash’s Amazonification of restaurants presents, which is “vertical integration” along the lines of the Amazon private label department that antitrust enforcers are currently investigating.

The Specter of Ghost Kitchens

“Dark kitchens” or “ghost kitchens” are anonymous commissary kitchens, often serviced by strategically located heating stations, that produce food exclusively for the delivery apps. A journalist touring one Los Angeles ghost kitchen found that it was selling its food through the apps under no fewer than 127 fake “virtual restaurant” names.  

UberEats, Grubhub and DoorDash all have busy “virtual restaurant” departments, with DoorDash’s dark kitchen executive boasting in an interview that the company’s superior and “highly specific” data on “hundreds of variables” regarding who eats what, when and where—CEO Tony Xu likes to boast that his company collects “hundreds of millions of data points”—could thoroughly “de-risk” the process of founding a new restaurant. 

Amazon’s private label division repeatedly and systematically used sales data to identify, copy and reproduce the most popular goods sold by its third party, ultimately bankrupting countless small businesses. But while Amazon has recently come under scrutiny over this practice, the rise of dark kitchens in the midst of an unprecedented restaurant industry bloodbath has attracted virtually no scrutiny. Former Grubhub head of innovation Collin Wallace tells Economic Liberties the combination of the pandemic and the ensuing rash of fee cap laws has ushered in a “dark kitchen land grab” he worries could spell doom for all but the most high-profile Michelin-caliber restaurants, and that the dark kitchen sector as it is currently constituted is rife with blatant “umpire player” conflicts of interest.

With this in mind, the delivery app cabal has worked hard to portray dark kitchens as an easy, low-risk way for chefs to open a “restaurant” without raising huge amounts of capital, with CloudKitchens most notably telling prospective partners that they only need $30,000 to start a “virtual restaurant” in one of their spaces. The reality for indie dark kitchens is much harsher: one study commissioned by a group of dark kitchen investors determined that when the labor costs and delivery app commissions were factored in, a dark kitchen needed to generate $675,000 in annual sales simply to break even. 

In lieu of culinary variety, delivery apps have sought out celebrity affiliations for their dark kitchens to give users the illusion of choice: UberEats has an exclusive virtual fried chicken chain modeled after the fictional methamphetamine front Los Pollos Hermanos in the television series “Breaking Bad”; GrubHub’s virtual fried chicken chain is called Tyga Bites, after the celebrity rapper. In the absence of national name recognition, Jim Collins, whose Pasadena-based company Kitchen United is one of the biggest dark kitchen operators, says he deliberately seeks out partnerships with established regional chains and shuns independent operators without multiple locations.

The companies that will succeed in dark kitchens will in all likelihood have deep pockets and financial connections to the delivery apps. The biggest dark kitchen startup in America, CloudKitchens, for example, was founded by Uber founder Travis Kalanick with some $700 million in cash and $2.1 billion in credit from investors including a Saudi sovereign wealth fund and Goldman Sachs. The country’s second fastest-growing dark kitchen chain, the Reef Kitchens, is a subsidiary of the parking lot chain ParkJockey and has hundreds of millions of dollars in funding from Softbank, the single biggest investor in both DoorDash and Uber, whose managing director Michael Ronen told the Financial Times in 2019 that the “success of UberEats [and] DoorDash” had led his firm to “try and stand up supply that is more efficient against that demand.” Kitchen United, got its seed investment and much of its Series A funding from Google Ventures, and Collins says he makes a conscious effort to “play nice” with delivery apps. The brother of the chief executive officer of Walmart’s e-commerce division recently launched another dark kitchen startup focused on using food trucks to prepare upscale big-city dishes for wealthy suburbanites. And Amazon itself is a substantial if quiet player in dark kitchens via its 16% stake in Deliveroo, a vertically integrated delivery app/dark kitchen conglomerate. Together with the in-house dark kitchen ventures run by DoorDash, Grubhub and UberEats, all the major dark kitchen startups have access not simply to vast pools of funding the restaurants don’t have, but data they don’t have—even though for the most part it was generated by them, and will now likely be used to copy and destroy their businesses.


Ensuring that restaurants and communities can connect through new technologies while preventing predatory practices by middlemen will require the deliberate structuring of markets by local, state, and/or Federal authorities. Not every city needs to adopt the same policy framework, but the principles of fair commerce, transparency, and no conflicts of interests will serve consumers, restaurants, and workers well. Below are nine policy ideas to aid in creating healthy restaurant markets.

1. Investigate and prosecute the apps’ systematic unfair and deceptive practices

Restaurants work hard to build trust and loyalty among their customers, and delivery apps capitalize on that relationship for profit while poisoning that trust. From creating phantom websites and fake phone numbers for partnered restaurants to using slightly altered logos and rampant flyering to give customers the illusion of complicity with unaffiliated restaurants, the delivery apps have built multibillion dollar businesses around deliberately misleading and deceiving consumers. The Federal Trade Commission and/or state attorneys general and local officials with jurisdiction should investigate and prosecute these harmful practices.

2. Prohibit delivery apps from imposing no price competition clauses.

“No Price Competition Clauses” prohibit restaurants from charging different prices across different platforms or even to on-premise or phone-in customers. The result has harmed both restaurants and their customers in two ways: First, it forces restaurants to penalize the on-premise consumers who are most profitable to the restaurant, by forcing those customers to pay the same price that Grubhub and UberEats users pay, despite the wildly different impact between either activity on the restaurant’s bottom line. Second, by removing the ability of the restaurant to differentiate its prices in line with the costs it sustains from using any of the apps and thus guaranteeing price uniformity across the internet, the clauses spare the apps most of the trouble of having to compete with one another’s prices. These clauses also have helped dramatically alter consumer behavior in a way that is detrimental to restaurants by not forcing consumers to internalize the cost of delivery.

3. Ban further anti-competitive mergers in the sector. 

Grubhub and DoorDash are together the agglomeration of more than 20 companies that were until recently independent, roughly three quarters of which did roughly the same thing. These mergers were executed to reduce competition within the industry and enable the apps to lower courier wages and raise prices without offering any improvement in service. Federal enforcers or state attorneys general should block any further consolidation and potentially look to reverse past mergers.

Regulators should keep in mind that the so-called “economies of scale” invariably invoked as the motivation for such consolidation are unusually elusive in the delivery app business, because both independent restaurants and “last mile” delivery derive little significant operational benefit from the national scale of a company like Grubhub or DoorDash. 

4. Enforce and expand local laws curbing predatory commissions and other delivery app abuses into broad state and local licensing regimes empowered to ensure compliance with all laws governing food service and alcohol sales.

In the wake of the COVID-19 pandemic dozens of cities, states and other municipalities have enacted emergency legislation capping delivery app commissions, mandating greater transparency regarding the fees associated with ordering over the apps, and—in Seattle—requiring the delivery apps to pay a $2.50 per order hazard surcharge to couriers. But the apps have broadly flouted the measures in cities and states where their lobbyists have detected loopholes in the language used and/or inadequate resources for enforcement. Sources close to the apps also say the delivery apps are preparing to sue many of the municipalities in question for violating their constitutional right to avoid uncompensated “takings” of the spoils of interstate commerce. 

One possibility is to establish licensing regimes for delivery apps, within the food service licensing authorities of county health departments and/or state liquor control boards. Local regulators can determine what commission structures and contracts are fair, impose fines and penalties on violators, and establish and enforce clear guidelines for contending with consumer complaints and refund demands. Just as critically, a licensing regime can ensure compliance with other laws routinely broken by the delivery apps, from improper collection of sales taxes to health department prohibitions on the consumption of food in a food handling environment. 

Local legislators, regulators and prosecutors should design regulation of the delivery apps with the intention of nurturing local services with a vested interest in the growth and support of local restaurants, specifically those that explicitly refuse to pursue national-scale strategies or venture capital financing, both of which lend little to delivery services’ marketing or operational effectiveness but typically invite predatory and anti-competitive practices. 

5. Prohibit delivery apps from using loss-leader pricing to harm competition and incentivize consumers to abandon on-premise dining.

From Amazon Prime to Uber and Lyft to DoorDash and Grubhub, deep-pocketed Wall Street and Silicon Valley-backed corporations have been subsidizing free and cheap “last mile” transportation services with the intent of driving out rivals and amassing market power. Until the 1980s, this business model was commonly understood to constitute “predatory pricing,” which violates statutes against anti-competitive and monopolistic behavior, because the number of companies that can afford to run losses for a sustained period of time is inherently limited. But predatory pricing doctrine has been substantially defanged by a series of narrow Supreme Court interpretations that force victims to meet onerously high burdens of proof, allowing predatory pricing to become an increasingly ubiquitous tactic among cash burning venture-backed startups from Lyft to WeWork. 

Proving predatory pricing has been hindered by the artificially high thresholds set forth in the 1993 Supreme Court decision Brooke Group v. Brown and Williamson Tobacco, which requires victims not only to prove that a company’s pricing is losing money for the company and that it is doing so with the express intent of running a competitor out of business, but that the company specifically will also recoup the losses at some point in the near future by raising prices. This ‘recoupment test’ presents an impossible burden for plaintiffs. Congress should amend this doctrine to bring the thresholds for proving predatory pricing in line with what is realistically provable by a competing firm—losing money for a sustained period on any specific product or service is inherently anti-competitive and should be illegal, period. 

6. Eliminate “independent contractor” loopholes and force the third party delivery giants to give their workers the wages, protections and benefits required of employers. 

Along with anti-competitive pricing, the business models of the third party delivery apps rely on their continuing ability to drastically under-compensate their workers by intentionally misclassifying them as “independent contractors.” In its 2018 decision Dynamex Operations West Inc. v Superior Court, the California Supreme Court established a strict and sensible three part litmus test for determining whether “contractors” were actually de facto employees, and in January 2020 the state legislature codified the guidelines, requiring companies like Uber and DoorDash to begin treating its drivers as employees, with the attendant stability, benefits and wage protections to which the law entitled them. But the third party delivery apps, along with the rideshare operators, chose to simply ignore the new law while spending tens of millions on lobbying and advertising in an effort to pass a ballot initiative carving out an exemption for themselves in November.

Allowing these companies to build their empires on the backs of illegally underpaid drivers unjustly penalizes not just the workers themselves but all companies that abide by labor law—including restaurants that employ, or have considered employing, their own delivery drivers. Congress should move to nationalize the litmus test established in Dynamex to protect both workers and law-abiding businesses that do not violate their rights in the name of “disruption.”

7. Require delivery apps to restrict the use of data collected from restaurants to  limited and specific purposes, and explicitly prohibit them from leveraging data  collected from transactions conducted over their marketplace platforms to  compete with and disadvantage restaurants.

Delivery app executives boast of collecting “hundreds of millions [of] points” of data from the restaurants on their platforms with virtually no restrictions on their use. Restaurants regularly complain that selling through delivery apps severs their relationships and ability to communicate with their customers. The apps, by their own admission, also use their data for another purpose: to develop “dark kitchens” in neighborhoods of high demand with menus specifically tailored to cannibalize the businesses of their existing restaurant partners. Regulators should issue guidance mandating that delivery apps share data they collect on restaurants with the relevant restaurant partners, and specifically restricting the use of their aggregate data to operations pertaining to their function as a neutral marketplace for restaurants. Such a ‘purpose limitation’ on the use of restaurant data comports with the Fair Information Practice Principles proposed by the U.S. government in 1973, forming the basis for such laws as the Privacy Act and the Fair Credit Reporting Act.

8. Mandate search neutrality within apps and bar payola style arrangements between apps and restaurants.

Delivery apps feature three types of restaurants on their platforms: partner restaurants that have signed contracts and paid a flat fee for access to the platform, partner restaurants that have (often unwittingly) agreed to pay for additional marketing and advertising support on those platforms, and restaurants with no relationship whatsoever with the delivery apps but from whom the apps have have chosen to offer delivery service to bolster the popularity of their marketplaces. Currently the formulas the apps use to determine which restaurants receive preferential exposure on their platforms are opaque and there is no way for either consumers or restaurants to determine what goes into a given restaurant’s search engine. 

The FTC should order the delivery apps to comply with guidances issued to generic search engines by clearly delineating between the three different tiers of inclusion on their platforms. Consumers have a right to know which restaurant listings are presented because of advertising or paid relationships and which are organic results conforming to the expectation that the app is delivering the most relevant result.

9. Separate platform and commerce in two ways: (1) Prohibit the combination of online ordering apps and delivery/logistics services (2) Online ordering apps and dark kitchens.

As Amazon private label’s systematic appropriation of product designs and concepts of its third-party seller demonstrates, there is an anti-competitive danger in allowing companies that own or control marketplaces of vendors to compete against those vendors using the superior proprietary information gleaned from their status as an industry “umpire.” Legislators should strictly prohibit third party delivery apps from owning or effectively controlling through contract arrangements “dark kitchens,” food trucks and other purveyors of prepared food that could unduly benefit from promotional or informational support.  


Running a restaurant is a difficult business. Nevertheless, over the past couple of decades it has lured in thousands of entrepreneurs because people have passion for making food and serving their communities.

Now those businesses are under siege—from a once-in-a-century public health calamity, but also a clique of monopolistic, rent-seeking predators who see them as just another vehicle for extracting dollars, and are exploiting the surge in demand for delivery services to raise more money, do more mergers, and roll out their own production facilities. 

The delivery apps have burned billions of dollars and broken dozens of laws with impunity while making it harder and harder for small restaurants to break even, and their vertical integration via the “dark kitchens” could displace small restaurants entirely. 

But while their tactics and strategies are cribbed from Google and Amazon, the delivery apps are not that powerful yet. There is still time to save America’s independent restaurants from going the way of our bookshops and toy stores. Our regulators, antitrust enforcers, and legislators must show the resolve to enforce our laws with all the determination and zeal with which the delivery apps have flouted them.



1  Elaine Underwood, “Postmates gathers 16 celebrities in support of local restaurants,” PRWeek, April 2, 2020.

2 Amelia Lucas, “ DoorDash eliminates and reduces some commission fees for restaurants,” CNBC, March 17, 2020.

3 Caleb Pershan, “Grubhub Asks Restaurants to Foot the Bill on ‘Supper for Support’ Promotions,” Eater, March 31, 2020.

4 Bruce Grindy, “Restaurant sales hit a pandemic high in June, but were far below normal” National Restaurant Association, July 16, 2020.

5 Eve Batey, “Delivery Apps Refuse to Temporarily Decrease the Fees They Charge Restaurants,” Eater, April 3, 2020.

6 Margaux Weeke, “City Council Unanimously Adopts Ordinance Limiting Third-Party Food Delivery Fees,” July 8, 2020.

7 Gabe Guarente, “Postmates Doesn’t Seem to Be Complying with a New Seattle Order Limiting Third-Party App Fees,” Eater, May 5, 2020.

8 Celina Tebor, “Restaurants, city say food delivery apps aren’t complying with Portland’s 10% commission cap,” Oregon Live, July 24, 2020.

9 Liyin Yeo, “Which company is winning the restaurant food delivery war?,” Second Measure, August 18, 2020.

10 Justin Phillips, “DoorDash admits to violating SF’s cap on restaurant delivery fees, promises refunds,” San Francisco Chronicle, July 10, 2020.

11 “If You Use Grubhub, Please Read–They Defied SF’s Delivery Commission Fee Cap,” Tablehopper, July 17, 2020.–they-defied-sfs-delivery-commission-fee-cap/

12 Underscoring the predatory nature of the delivery apps, complaints over food temperature, long waits and inaccurate menu descriptions typically result in consumer refunds that are systematically charged back to the restaurants that exercise little control over those things. Anna Schmidt, “How restaurants, food delivery apps like Grubhub, Uber Eats frustrate consumers: Survey,” FOXBusiness, January 8, 2020.

13 While the rise of the delivery apps has also had a dramatic impact—and often a negative one, as lawsuits against the apps filed in recent years by In-N-Out Burger, Yum Brands and Legal Sea Foods will attest—on chain restaurants, Economic Liberties has deliberately chosen to focus on independent restaurants for the purposes of this paper. Chain restaurants have by and large used their size and marketing budgets to negotiate far superior terms from delivery apps—typically along the lines of 15%—than independent restaurants have the leverage to, meaning the breakeven business McDonald’s and Popeye’s do with the apps is typically subsidized by smaller operators, which typically generate an average $4 per order in profit for the platforms, as this excellent story explains in greater depth. See Adrianne Jeffries, “During the Pandemic, Grubhub Should Be Thriving. It’s Not,” The Markup, May 27, 2020.

14 Matt and Cathy Bussewitz, “Uber buys Postmates, ups delivery game in $2.65 billion deal,” ABC News,  July 6, 2020.

15 Erin Griffith, “DoorDash Faces Its Latest Challenge: Wooing Wall Street,” The New York Times, May 4, 2020.,because%20the%20information%20was%20private.

16 Sherin Shibu, “One of the most successful investors in Silicon Valley says hearing these 2 things will instantly make him love a startup pitch,” Business Insider, July 8, 2019.

17 Billy Gallagher, “DoorDash Raises $2.4M To Improve Food Delivery,” September 30, 2013.

18 Arash Massoudi, Eric Platt, James Fontanella-Khan, Miles Kruppa, “Uber and Doordash held merger talks after Softbank push,” Financial Times, January 30, 2020.

19 “RETAILPostmates Notches $100M In Funding,” PYMNTS, January 11, 2019.

20 Grubhub’s financial statements suggest average commissions have risen from just over 14% in 2014 to 22%. It is offering more of its customers delivery services, but also doing more volume on behalf of large chains that command steep discounts in their commissions.

21 An analysis of 229 November 2019 and December 2019 DoorDash driver pay reports conducted by the state labor rights advocay group Working Washington determined that driver pay after calculating for mileage and additional payroll taxes averaged $1.45 per hour. The survey is available at

22 Erin Wade, “An Open Letter to Matt Maloney, CEO of Grubhub,

23 Postmates, Grubhub and UberEats all specifically bind restaurant partners to charge the same prices across the board in their default contracts. DoorDash has been somewhat more flexible about pricing but informally enforces the same policy, according to industry experts and attorneys. See Davitashvili v GrubHub Inc., 20-cv-3000, U.S. District Court, Southern District of New York. Charges Sales Tax on Orders in Tax-Free States,”, August 5, 2019.

24  Andy Newman, “My Frantic Life as a Cab-Dodging, Tip Chasing Food App Deliveryman” The New York Times, July 21, 2019.

25 The extent to which DoorDash couriers are aggressively supervised throughout a given shift is detailed in San Francisco District Attorney Chesa Boudin’s lawsuit against DoorDash. See The People of the State of California v DoorDash Inc., CGC-20-584789, U.S. Superior Court, City and County of San Francisco  

26 In 2017 the FBI launched an investigation into a software program code-named “Hell” devised by Uber to spy on drivers it suspected of “double-apping” as Lyft drivers in an attempt to coax those drivers into promising exclusivity. See Rebecca O’Brien and Greg Bensinger,“Uber Faces FBI Probe Over Program Targeting Rival Lyft,” The Wall Street Journal, September 8, 2017.  

27Yes on 22, “Key Facts about Proposition 22,”

28What is the ‘Gig Economy’ and Who Works In It?, Countable, October 24, 2019.

29Harry Campbell, “Lyft & Uber Driver Survey 2019: Uber Driver Satisfaction Takes a Big Hit,”August 1, 2020.

30 No free lunch, but almost:, PayUp, July 16, 2020.

31 DoorDash bowed to pressure to change its tipping policy after it was detailed in this New York Times story, but especially during daytime hours tips are still factored in to “guaranteed minimum” rates; the company is simply more forthcoming about the policy with drivers. See Andy Newman, “My Frantic Life as a Cab-Dodging, Tip Chasing Food App Deliveryman” The New York Times, July 21, 2019.

32DoorDash charges workers $15 for “free” coronavirus supplies — while paying them $2 to deliver food, PayUp, March 14, 2020.

33 “In California, Doordashers sue over gig worker classification,” The Counter, February 13, 2020.

34Andrew J. Hawkins, “Uber, Lyft, and DoorDash kick off $90 million fight against California’s gig worker law,” The Verge, October 29, 2019. So far, together with Instacart, the apps have spent $110 million promoting Proposition 22. See Sarah Holder, “California’s Gig Economy Ballot Measure Fails Workers, Labor Groups Say,” July 7, 2020.

35 New York, Dallas, Houston, Miami and Philadelphia are more than 50% controlled by Grubhub, Doordash, Doordash, UberEats and Doordash respectively. See Liyin Yeo, “Which company is winning the restaurant food delivery war?,” Second Measure, August 18, 2020.

36Joshua Franklin and Anirban Sen, “Postmates revives IPO plans amid takeover interest,” News Break, June 30, 2020.

37  Shirin Ghaffary, “DoorDash and Uber Eats aren’t collecting sales tax on delivery fees in some states. That could be a problem.” Vox, November 20, 2019.

38 A lawsuit seeking class action status to recoup the improperly charged sales taxes was settled out of court after DoorDash attempted to enforce a mandatory arbitration clause. Notably, the complaint noted that DoorDash had not improperly charged customers sales taxes in Oregon or Alaska. See Erin Shaak, “Class Action Claims DoorDash Charges Sales Tax on Orders in Tax-Free States,”, August 5, 2019.

39Claire Brown, “GrubHub is buying up thousands of restaurant web addresses. That means Mom and Pop can’t own their slice of the internet,”, June 28, 2019. Grubhub and its subsidiary Seamless were not the only delivery apps to do this; in 2015 another investigation found that the then-upstart delivery app OrderAhead, which is now owned by DoorDash, acquired several thousands of restaurant URLs. While both Postmates and DoorDash pioneered their own versions of the same strategy, even making it more abusive in many cases, the annexation of restaurants’ Google presences is most closely associated with Grubhub.

40Federal Trade Commission, “Sample Letter to General Purpose Search Engines,” June 24, 2013.

41 Victor Fiorillo, “Grubhub Hit With Huge Class-Action Suit by Philly Indian Chain Tiffin,” Philadelphia magazine, January 4, 2019.

42“Promotions” were the culprit for jacking up a Portland vegan restaurant’s total Grubhub tab to 65%, a Chicago pizzeria’s Grubhub tab for the month of April to 64% and a Pittsburgh diner’s April Grubhub tab to 58%. See Tim Forster, “Grubhub and Postmates Are Actively Defying Portland’s New Delivery Fee Law,” Eater Portland, July 29, 2020 and Nick Kindelsperger, “Chicago restaurant chose the extras, Grubhub says,” Chicago Tribune, May 1, 2020 and also Angie Moreschi, “Grubhub marketing fees leave Pittsburgh restaurant owner shocked, upset,”, May 18, 2020.

43 Eliza Fisher, “What restaurateurs need to know about third-party delivery services and Google ‘Hijacking’,” UsFoods.

44 The private equity consortium had taken control of Seamless via a 2006 leveraged buyout of the Philadelphia institutional food provider Aramark, which owned Seamless and spun it off to its owners as a dividend in 2012. The PE group acquired roughly 65% of Grubhub’s shares in 2013 in what at the time was billed as a merger of equals between Grubhub and Seamless, though they allowed Grubhub CEO Matt Maloney to keep the CEO title. See Telis Demos and James MacMillan, “GrubHub Seamless Joins the Tech IPO Crush,” Wall Street Journal, February 20, 2014. 

45 Katherine Carlon, “Squeezed by big guys, CHOMP bites back,” Corridor Business Journal, October 30, 2017.

46 Calculated by dividing revenues by gross food sales reported in Grubhub’s 2013 vs. 2018 fiscal years: 

47 CNBC, “Grubhub CEO Matt Maloney on coronavirus, its new membership program and more,” February 26, 2020.

48 Aileen Lee, “Welcome to the Unicorn Club: Learning from Billion Dollar Startups,” TechCrunch, November 2, 2013.

49 According to Crunchbase, DoorDash has raised $2.5 billion in venture funding as a private company, while Grubhub has raised a mere $284 million inclusive of its public stock offerings—and only ever raised $85 million in the five rounds of funding it did as a private company. 

50 The company articulated its original business idea in 2013 as “a local, on-demand Fedex” in a 2013 Medium post But Fedex, founded in 1971 by the heir to a restaurant and bus company fortune, was the product of a vastly different generation of venture capital economics and was turning a healthy profit by 1974, despite start-up costs that included a fleet of planes and a full-time staff of pilots who did not pay for their own fuel.

51 Grubhub did not offer its own delivery services until around 2014, when the emergence of Doordash inspired it to roll out its own delivery network in spite of its leaders’ belief it was a “shitty business.” See Brian Solomon, “Why GrubHub Is Building What Its CEO Calls ‘A S***** Business’,” Forbes, April 20, 2016.

52 DoorDash controlled 33% of the American food delivery market in 2019, according to the research group Second Measure. Although it has yet to turn a profit, the company has burned cash effectively enough to secure itself 11 rounds of venture capital cash worth $2.5 billion. See CrunchBase 

53 See Legal Sea Foods v. DoorDash Inc. and In N Out Burgers v. DoorDash Inc. In 2018 the company was also the subject of a similar lawsuit filed by the Chicago chain Burger Antics, see Robert Channick, “Suburban restaurant in DoorDash Lawsuit: Stop Delivering our food”, Chicago Tribune, January 10, 2018. 

54 In fact, as your analyst was composing this report her husband, the chef of a hotel restaurant that has been closed for two months, received a phone call from a concierge reporting that a courier representing a delivery app he refused to name had arrived to pick up a pizza and chicken entree that had been ordered over one of the platforms, none of which he had ever entered into a contract.

55 Matt Newberg, “My Testimony for the Nation’s First Hearing on Ghost Kitchens,” Medium, February 6, 2020.

56 “Data at DoorDash: Transparent, Ubiquitous, and Still Just Getting Started,” Medium, October 21, 2016.

57 Sara Salinas and John Fortt, “DoorDash CEO Tony Xu: ‘We need to know about about every pothole and parking space to perfect our food delivery,” CNBC, March 24, 2018.

58 Peter Romeo, “How DoorDash’s Ghost Kitchen Works,” Restaurant Business, October 16, 2019.

59 Dana Mattioli, “Amazon Scooped Up Data From Its Own Sellers to Launch Competing Products,” Wall Street Journal, April 23, 2020.


 Michelin-starred restaurants aren’t immune from the consequences of dark kitchen disruption, however; the Michelin-starred Thai restaurant Kin Khao briefly had its identity stolen from a food truck based dark kitchen owned by the Softbank-funded Reef Kitchens. See Chris Hua, “Softbank funded the Kin Khao Impersonator,” “Optimism of the Will”  SubStack, January 27, 2020.

61 Samantha Leffler, “Food From Breaking Bad’s Los Pollos Hermanos Chain Is Now Available Via Uber Eats: ’No Way!’” Us Magazine, October 29, 2019. 

62“Tyga Opening Virtual Chicken Restaurants,” TMZ, July 29, 2020. 

63 Jennifer Marston, “Uber Eats Is doing Ghost Kitchens: Here’s How That Could Change Food Delivery,” The Spoon, March 12, 2019.

64 Tim Bradshaw, “The start-ups building ‘dark kitchens’ for Uber Eats and Deliveroo,“ Financial Times, May 21, 2019. 

65 I-Chun Chen, “Google-backed Kitchen United raises $40 million Los Angeles Business Journal, September 19, 2019.

66 Connie Loizos , “In the heated cloud kitchens race, Kitchen United aims to kill with kindness; here’s its playbook,” TechCrunch, November 19, 2019.

67 Matt Newburg, “Walmart Exec Bets Big on Suburbs in Stealth Ghost Kitchen Startup,” Medium, February 26, 2020.

68 Amazon shut down its own restaurant delivery service the month after announcing its Deliveroo investment, citing the fierce competition of the delivery apps but probably in response also to the concerns of European antitrust regulators.
See Sam Shead, “Amazon is one step closer to taking a 16% stake in UK delivery app Deliveroo,” CNBC, June 24. 2020