At the end of 2019, the U. S. restaurant industry was relatively stable with $900B in revenues, and 15.6M employees in the United States.1 Its composition included everything from fast food to Michelin 3-star restaurants, and from mom-and-pop local eateries to multi-national conglomerates. However, Covid-19 has wreaked havoc on the restaurant industry with losses of $240B by the end of 20202 and significant questions on its long-term sustainability.

Download the full PDF.


Even before the pandemic shuttered in-store dining, a growing portion of restaurant business was shifting to off-premise.3 Last year alone, 60% (or $540B) of food service occasions were pick up, drive-thru, or delivered. And these are not just quick-service restaurants, like McDonald’s. With the proliferation of third-party delivery services such as Grubhub, DoorDash, and UberEATS, the food aggregator market is expected to grow rapidly from $10B in 2018 to $75B by 2022.4

The growth of disruptive business models promises to shift the basis of competition for restaurants, and COVID-19 is only accelerating this shift by reshaping consumer behaviors. We’ll discuss how three driving forces will reshape the future of the food industry and explore which types of restaurants are most vulnerable.


The invention and meteoric rise of digital aggregators have reduced customers’ friction for discovering restaurants, exploring cuisines, and accessing personalized options at their own convenience. As dramatically as this market has grown, aggregator business models are not likely to disrupt the industry in the classic sense. These business models essentially add a new channel, but fundamentally operate within the same economic construct. The extra cost of food delivery has been largely passed on to the consumer and vendor.

More likely to disrupt the food and restaurant industry are “dark kitchens.” These are lean spaces that are dedicated exclusively to food delivery and takeout.

More likely to disrupt the food and restaurant industry, however, is the rise of “dark kitchens.” These are lean spaces that are dedicated exclusively to food delivery and takeout. A popular version is a shared commercial space that holds up to 20 kitchens and is leased to different food operators. This model offers unique quality trade-offs to consumers with lower price and greater choice and is designed to operate through a new economic model for businesses, with lower capital expense, reduced delivery cost, and higher labor utilization.

The concept has caught the attention of Silicon Valley, where investors are putting hundreds of millions of dollars into dark kitchen startups. CloudKitchens was launched by Travis Kalanick (of Uber fame), Virtual Kitchen got a sizeable investment from PayPal founder Peter Thiel, and Kitchen United has been backed by GoogleVentures.5 They use in-house data, advanced analytics, and rapid testing to optimize floor space, delivery patterns, and consumer palate customization, a significant shift from how the industry has historically operated.

Since the infrastructure for dark kitchens is turn-key, it allows a “digital restaurant” to launch in a matter of weeks rather than months. Capital expenditures are either low, or non-existent, and average lease costs are below standard brick-and-mortar establishments, particularly because underlying real estate is optimized for off-main street urban locations. With no front-of-house, labor costs are noticeably leaner. Aside from cost-structure improvements, dark kitchens often provide members access to data hubs for standardizing costs and prices, understanding consumer preferences, and utilizing forecasting algorithms for better supply management. Dark kitchens could also solve the “last-mile” problem as couriers can pick-up several different restaurant orders at once with greater efficiency.

Finally, dark kitchens can fully digitize a restaurant’s brand presence. With no menus to print, no analog signs to change, and no established décor, a restaurateur can iterate through hundreds of changes in a matter of weeks, allowing for digital A and B testing that was previously reserved for digital-only products. Further yet, a restaurateur can add virtual brand spin-offs out of the same dark-kitchen with minimal cost, but potentially significant market-share gains. Figure 1 below shows the spectrum of operational and real estate control that the brand has in various dark kitchen options:




The increased demand for off-premise food begs the question: What purpose does a brick-and-mortar restaurant really serve? The theory of “jobs to be done,” introduced by Innosight cofounder Clay Christensen provides an important lens for helping answer this question. This theory explains that consumers don’t shop for products but rather seek solutions to make progress on the everyday “jobs” they encounter in a given circumstance.

With this in mind, a traditional restaurant has historically been a good solution for three primary customer jobs: the functional job of food preparation, the social job of socializing with friends and loved ones, and the location-related job of dining with an ocean-front view or being able to access a specific venue with limited food options like a sports stadium or an airplane.

The jobs lens is helpful in developing a broader view of any industry’s competition. For example, it explains why the Thai restaurant down the street competes not just with the ramen shop around the corner, but with a Netflix subscription and frozen noodles from Trader Joe’s. Arguably an evening home with your desired company, binge-watching the latest season of “Stranger Things” while microwaving your dinner, addresses similar jobs around food preparation, entertainment, and social interaction, as the Thai restaurant otherwise would.

A recent Morgan Stanley study found that 43% of consumers who ordered food for delivery had actually replaced it with a dine-in meal at a restaurant.

Increased competition for consumer jobs from non-restaurant sources, such a Netflix and dark kitchens, have called into question what was previously considered restaurant mainstays. With these pressures, historically razor-thin industry margins, and demanding consumer expectations, many restaurants looked at delivery aggregators like Grubhub as a way to grow volume. The hope was that the lower marginal cost of an additional meal would offset the 15% to 30% commission that delivery apps charged and still keep restaurants vying for in-home meal replacements.6 However, a recent Morgan Stanley study found that 43% of consumers who ordered food for delivery had actually replaced it with a dine-in meal at a restaurant.7 This uneven exchange simply shifted control to technology companies, whose platforms doubled as marketing and discovery hubs.8

Even international giants such as McDonald’s and Subway, which have traditionally relied on their brand equity, became increasingly dependent on these platforms for delivery logistics. In response, since physical presence and foot traffic were often their differentiating factors, many restaurants doubled down on longer-term capital leases to compete on cost advantages and greater scale. This response is not uncommon for large incumbents operating in industries undergoing disruption. However, you cannot beat disruption by simply scaling or optimizing a challenged model, you need to develop a fundamentally different way to compete.

That is certainly what the co-author of this article, Sam Pogosov, discovered after starting Phinix Mediterranean five years ago in a Boston suburb. It took him six months and hundreds of thousands of dollars to open his first storefront – a traditional restaurant with standard features such as dine-in and take out throughout the day. Three years later, while revenues had doubled year over year, increased labor, rent, food waste, and built-in inefficiencies kept his net margins in the low single digits.

Faced with another sizable lease extension, Pogosov decided to “go dark,” shutter his store-front, and scale down to run his brand out of a new 800-square-foot kitchen. Today Phinix has no seating or storefront, and is instead focused on churning out orders for delivery across a variety of brands. To the casual customer, Phinix is as readily available on UberEATS as a physically present Chipotle or McDonald’s, while its operating costs have decreased by over 70%.

Other notable incumbents are testing the waters of the model too. Ruby Tuesday has started providing host kitchens that allow outside brands to use their excess kitchen capacity strictly for delivery.9 McDonalds is testing franchise expansion through dark kitchens in the UK, and large hospitality groups such as SBE are making significant shifts in providing high-end delivery-only brands. All of this activity signals a fundamentally new and different view of the future competitive landscape.


Powerful disruptive events such as the pandemic can create the motivation to modify entrenched habits and “unfreeze” behaviors for a period of time.

While disruptive forces, transformative new models, and shifting competition have been reshaping the food industry for some time, the obvious major accelerant has been COVID-19. Government shut-downs turned virtually every restaurant that didn’t close into a pick-up and delivery-only service, with 92% of restaurant traffic shifting off-premise overnight. Consumer adoption of digital ordering and payment platforms grew at a staggering pace, with 51% of consumers downloading at least one new food app in May 2020.10 The key question remains, what new consumer behaviors will get forged during the crises and how will they determine future expectations of experience?

Psychologist Kurt Lewin’s “change model” proposes that individual behaviors are typically “frozen” and hard to change. However, powerful disruptive events such as the pandemic can create the motivation to modify these entrenched habits and “unfreeze” behaviors for a period of time. This leads to a window of experimentation when we transition to a new set of behaviors, habits, and affiliations. This is where we are now.

As a result, organizations should seek to understand if consumers will learn whether a temporary job to be done dislocation, behavior, or solution is better or worse than what they were previously accustomed to. That will indicate the likelihood of a behavior “refreezing” post-pandemic. For example, while virtual schooling has been a temporary compensating behavior, the experience of parents suggests that the majority of children will likely return to school when the pandemic subsides. By contrast, many industries have witnessed similar levels of work productivity at lower costs with the home office model, leading organizations like Twitter and Square to make it a more permanent option in the future.11

While restaurants are slowly reopening, capacity limitations, increased health precautions, and higher operating costs are exerting greater financial pressures. More than 16,000 have permanently shut down.12 Consumers are demanding a new age of food discovery, one with contact-less delivery and hyper personalization, and where the “convenience” of home delivery is a table stakes expectation and not a premium-priced differentiated service. As a whole new generation of consumers gets accustomed to these new behaviors, industry incumbents will have to either transform their business model, further differentiate beyond food, or risk becoming a commodity offering at the mercy of digital food aggregators.


The three primary forces highlighted above are rapidly changing the future of the food industry. To help incumbents better assess their future sustainability, we have developed the “Restaurant Basis of Competition Model” in Figure 2 below. This model disaggregates the three primary customer jobs-to-be-done different types of restaurants have historically competed for and highlights what types of food service organizations are most vulnerable to disruption.



High Threat of Disruption (A): Fast food, Mom and Pop stores and other proximity-based restaurants are likely the most vulnerable to disruption. These restaurants focus on fulfilling functional jobs, that is, access to prepared food, and represent the simplest form of food preparation. They can be brick-and-mortar walk-in businesses, have drive-thrus, or even operate as delivery-only (such as pizza delivery). Since they offer basic services, there is little room for differentiation.

It is important to isolate desires by customer’s primary Desire, so while many restaurants will want to move beyond this circle, a non-differentiated, walk-in brick-and-mortar business in a big city that typically wins on the convenience of physical proximity, is most vulnerable to disruption.

Medium Threat of Disruption (B): Restaurant chains like the Capital Grille, Dave and Buster’s and other pubs and bars focus on fulfilling social and emotional jobs to be done. These restaurants create a social ambience through either direct social cues such as game machines or indirect prestige of high-end dining (like Capital Grille). In either case, customers are drawn almost exclusively for the social or emotional multipliers, though food is still a key component. A good proxy to differentiate between this social and food desire, would be to evaluate a percentage of “to-go” food orders. If a majority of the dining and activity are on-site, the establishment likely caters to social desires.

Low Threat of Disruption (C): Sports arenas, ocean-front casual dining, and theme parks focus on fulfilling location-based jobs to be done and use their location to drive customer volume, while food is an ancillary service at best. A simple testing criterion would be to get rid of the food-component of these types of restaurants, and attendance to the location should not change significantly. This also means that rural restaurants where the scale of competitors is minimal or non-existent have a natural monopoly based on location limitations.

The overlapping circles are hybrid cases and these generally have a couple of different strategic options which can be analyzed through further exploration.

Food preparation and service is the second most common occupation in the United States, with waiting tables being the eighth most common.13 A UBS study estimates that 1 in 5 restaurants will permanently close due to the pandemic.14 Those that survive will find themselves playing in a significantly redesigned competitive landscape, one that requires fundamentally different capabilities and economic models to win. This is not the time for the restaurant industry to simply try and survive the crisis. A new age of competition is emerging in the historically fragmented restaurant industry, where disruptive new entrants rely on data and digital technologies to create a new baseline of food and offer personalized experiences. With a more proactive approach to understanding consumer behavior, and developing novel business models, savvy incumbents may not just survive but thrive in the post-COVID era.


About the Authors


Prashant SrivastavaPrashant Srivastava is a partner at Innosight.





Sam Pogosov the founder of Phinix Virtual Kitchens Group – a dark kitchen specializing in healthful brands and a former consultant at Innosight.




Yury Adamov is a strategy and business development manager at WarnerMedia and adjunct professor at Pepperdine University.