Third-party restaurant delivery is a major consumer innovation that links ordering through marketplaces to fulfillment by gig drivers managed by marketplace algorithms. It began in fits and starts — mostly just ordering marketplaces, with restaurants left to figure out fulfillment on their own. Third-party restaurant delivery accelerated when DoorDash linked the front and back ends together, effectively surrounding the restaurant in what initially appeared to be services the restaurant could not do on its own.
Third-party delivery has, in many ways, been a boon to all involved. Although not all delivery sales are incremental to on-premises sales, restaurants have generated incremental sales. Consumers have greater choice and convenience in procuring prepared meals. Millions of couriers collect extra income from The Gig ATM, effectively monetizing their auto assets by delivering restaurant food. The marketplaces make it easy for restaurants to get online, make it easy for consumers to order from a wide choice of restaurants, and make it easy for workers to find gigs. These platforms have successfully brought together three different stakeholders in a way unlikely to have occurred without them.
While the primary allure to all of these stakeholders is “easy to,” that upfront ease obscures downstream friction resulting from third-party platforms. Assuming that restaurants have figured out how to price for marketplace commissions and that delivery consumers have determined that paying higher prices for convenience is worth it, five major problems have emerged for restaurants now that third-party platforms are ubiquitous.
Challenge #1: Overwhelmed Kitchens
The idea of incremental sales is alluring. All restaurateurs know that transactions are the lifeblood of each restaurant location. More transactions are almost always better — except when they all occur at the exact same time.
As delivery has become a more ingrained behavior, it has become clear that most delivery orders occur simultaneously with most on-premises orders. This has left restaurants in the unusual position of having their kitchens controlled entirely by another business — the third-party marketplaces.
Consumers choose what to order based more on the menu than anything else. If you are a brunch place, expect morning orders to peak at the same time as your dine-in traffic. If you are a lunch place, expect lunchtime orders to peak at the same time as your dine-in traffic. If you are a dinner place, expect evening orders to peak at the same time as your dine-in.
Challenge #2: Food Waiting for Driver Arrival
“Get ahead of the wave” has been said in most kitchens. “Always be baking” is another similar phrase often used in pizza restaurants. The idea is that you know the peak is coming, so line staff should do whatever they can to get ahead. This might mean prepping ingredients, cooking 100 burgers knowing that someone will order them, or dropping in fries even if there are no orders on the KDS.
What this mindset has come to mean in terms of delivery is that if a delivery order comes in and there is capacity to make it, then make it. Make it without knowing where the driver is or when they will arrive. Make the difficult or capacity-constrained items first, even if they will degrade while the team pulls together the other items. Just make it.
While this mindset, in theory, increases capacity utilization by freeing up the team to make another order later, it often results in food sitting and waiting for the driver's arrival. Nothing good happens to food while it sits, no matter how good the packaging is. And the more food sits, the more shelves are necessary. And the more shelves a restaurant is filled with, the more it looks like an e-commerce fulfillment center and not a pleasant dining experience. The more e-commerce fulfillment takes over, the more correct order-to-driver matching becomes a problem.
Challenge #3: Restaurants Crowded with Non-Customers
Restaurants with more than 20% of their sales coming from delivery put their assets and teams under significant stress. With one in five parties present merely to collect an order on behalf of someone else, that means one in five parking spots are not directly generating revenue, one in five cars in the lot are blocking parking for on-premises guests, and one in five people trying to talk to the hostess merely want to collect the right bag.
Regardless of the professionalism, or lack thereof, of any given courier, these additional people can crowd the restaurant, again at peak time, causing both asset and front-of-house staff to work harder than the model was designed to do. This problem is exacerbated by the lack of coordination between courier arrival and food prep. While many restaurants try to “get ahead” of delivery orders, the courier often arrives before the food is finished. This wait time — frustrating to restaurants, couriers, and consumers alike — reduces parking spot turnover and furthers crowds in on-premise pickup areas.
Challenge #4: Chargebacks and Fraud
Recent surveys of restaurant owners have found that a shocking 3-4% of delivery sales are charged back to the restaurant. When delivery makes up 30% of sales, this can translate to more than a point of sales missing from the restaurant. Worse, the food has likely gone out the door. When food and paper costs are 30% of sales, restaurants could lose a half margin point just to chargebacks.
Perhaps the restaurant made a mistake, perhaps the courier consumed the food, or perhaps the delivery guest committed fraud by reporting an item or order missing that was, in fact, received. In most cases, it is impossible to say where the problem occurred. But marketplaces generally put the liability on the restaurant: guilty until proven innocent.
Delivery is not necessarily more prone to errors or fraud than is dine-in. The fragmented system just makes it harder to detect and correct errors and fraud. Where did the problem occur? Who is responsible? How can the problem be remedied? Can it be remedied?
While the guest relations and profit problems associated with chargebacks and fraud are obvious, the operational costs are less obvious. Owners, managers, and finance departments across the country incur these costs by investigating and disputing each individual chargeback.
Challenge #5: Lack of Guest Relationship
The last problem may be the most nefarious because its impacts are difficult to immediately quantify. Restaurants have lost their relationship with their guests and, with it, their ability to engage in hospitality.
Many restaurants lament the lack of guest data coming through third-party channels. Who is ordering? How often do they order? What do they order? While this lack of data seriously affects feedback loops and a restaurant’s ability to remarket to the guest, it isn’t really that different from recent history.
As little as 10 years ago, few restaurants knew the answers to these questions at scale. When sales were down at a major national brand, the Consumer Insights department deployed massive surveys to determine why- no behavioral data told the restaurant whether the problem lay in frequency or penetration. Were fewer customers coming in? Or were the same customers coming in, just less often? No one knew for sure. And yes, having this data is better than not having this data, but something else has been lost as a third party has been inserted between the guest and the restaurant.
The real problem for the restaurant is its complete inability to engage with the end consumer. The restaurant has no control over who represents its brand or how its brand is represented through couriers. The restaurant has no control over batching several deliveries into a single run. The restaurant has no control over fixing an issue when it arises.
Couriers are people. And like all people, some are more professional, and some are less. A specific restaurant may have a standard in mind for their employee brand representatives, but with a third-party system, there is no way to enforce this standard. The restaurant gets what it gets, in accordance with the invisible hand of the algorithm.
Some couriers engage in “self-batching,” combining orders from different restaurants to different destinations by using multiple phones to simultaneously deliver for multiple platforms. These couriers can increase their earnings by taking two orders from similar parts of town. Anyone who has ordered from a marketplace and watched their courier’s car stop at one or two other places of business, then one or two other places of consumption, knows how maddening this is. Delivery times run longer, food degrades further, and the consumer regrets the tip they placed at order, which no longer seems justified.
In response to this behavior, third-party platforms did not see a problem that needed to be eliminated. Instead, the marketplaces saw an opportunity to capitalize on. Third-party platforms now algorithmically batch orders. While this may be marginally better than the judgment of a courier running multiple phones, it’s still not great. Even well-batched orders will be slightly delayed, resulting in degradation in food.
As YouTube influencer Mrwhosetheboss pointed out in his recent video “The Internet Is Starting to Break: Here’s Why,” platforms use this degradation in service to not only enrich themselves (two delivery and service fees for one delivery) but also to create tiers of service for which they can further upcharge consumers. Don’t want to wait on the risk that the platform batches your order with one from across town? Pay an extra $2.99 for expedited service.
And where is the restaurant in all of this? Odds are the consumer will receive their order of cold, old food and blame the restaurant. Every front-of-house employee is trained in guest winback in case something goes wrong in a transaction. But how can a restaurant employee win back a guest who is unknown to the restaurant and complains to a third party? Worse, what if the complaint is a low star rating and an essay about the quality of the food on the platform, when that quality issue was created by the platform itself?
Getting Rid of Third-Party Complications
Over the last 10 years, a huge amount of innovation has gone into the consumer-facing front end of these challenges. If only we could get consumers to order directly from the restaurant, the thinking goes, all of these problems would go away. First-party online ordering systems have improved dramatically over the last 10 years, with many systems rivaling the frictionless ease of ordering through a marketplace.
Changing consumer behavior is harder than it seems, and merely offering consumers the ability to order directly from a restaurant has not always resulted in consumers ordering directly. Enter loyalty programs. Loyalty programs purport to give consumers enough value from ordering directly that they are more likely to do so. Loyalty programs have likewise come a long way, now offering consumers benefits beyond mere discounts and collecting behavioral data while they do so.
But still, the third-party marketplaces persist. As long as they offer restaurants access to consumers they might not otherwise get, consumers convenience that takes memorized passwords and credit card numbers to replicate, and drivers the ability to earn money flexibly that they might not otherwise make, the third parties will continue to be a part of running restaurants.
Solutions to Date Are a Start, but Inherently Reactive
Every Venture Capitalist loves a good problem statement. A great product becomes irresistible to its target market because it solves a genuine problem, a pain point. And the interesting thing about restaurant food delivery in 2024 is that it creates SO. MANY. PROBLEMS. These problems create rich soil for technological innovations to grow in.
Any restaurateur engaged in delivery receives many emails, phone calls, and visits each day from well-meaning sales reps of startups offering point solutions to each of these many problems. The solutions are amazing.
And dizzying. There are cameras and scales to reduce chargebacks. There are shelves, lockers, and packaging to deal with food waiting for drivers. There is throttling software and dynamic pricing to eliminate super-peaky demand. There are tools to select delivery fulfillment companies, geofence the restaurant, automate chargeback dispute resolution, and automatically ensure a restaurant’s online store on the marketplace is up.
Getting to Root Cause Prevents Problems by Solving from First Principles
But what if none of these tools were necessary? What if the solution to the problem is not having the problem in the first place? Is there a way to give consumers what they want, without creating a disjointed system that results in so many problems?
Designing from first principles means stepping back to determine what we are trying to achieve (great food delivered cost-effectively to a consumer). Instead of adding incremental solutions to an existing model, first-principles design solves for the objective, not the problems, by creating a well-coordinated, purpose-built system specifically for that objective.
The magic of a well-coordinated system is that everybody wins. With increased efficiency, the restaurant, the driver, and the consumer are all better off. While a well-coordinated system may take more effort upfront to set up, it is easier and more profitable once it is in place.
Does this, therefore, enable new entrants to the restaurant industry to have an advantage over current incumbents? Maybe. If we ask the executives of top restaurant chains what they would change about their operating system, many would likely touch on specific functions that need improvement. We would argue that it's not the functionality that should be considered initially but the mindset of those shaping the operating system that needs to change. A new entrant can design from a blank sheet. An incumbent must build on what is already generating revenue, so it is understandably not easy.
Perhaps it should not surprise us that many incumbents are designing stand-alone digital kitchens and their respective operating systems, intending only to service off-premise orders. As we discuss in the final chapters of Delivering the Digital Restaurant: The Path to Digital Maturity, restaurant owners will reach a fork in the road on how they decide to handle the challenges of food delivery. Some may be tempted to lean deeper into experiential dining and forgo the incrementality of delivery. Others may need to develop stand-alone units with systems to optimize their delivery experience - something we believe can only be achieved through a vertically integrated tech stack.
If you are tired of your guests complaining about your delivered food being old, cold, or missing….
If you are tired of a crowded parking lot and bags of food lined up where guests should be….
If you are tired of sorting through chargebacks to determine what is your restaurant’s fault….
If you are tired of your staff not being sure which order to prioritize during peak….
Maybe it’s time to look at how your multi-channel restaurant needs a fresh perspective. One that doesn’t attempt to solve the problems that come from delivery but instead tries to do delivery well. Something that gives your guests as good an experience through delivery as they may have sitting inside your restaurant itself.
As we have written in “Delivering the Digital Restaurant: Your Roadmap to the Future of Food” in “Chapter 5: Why Pizza Works,” there is a way to deliver without all the headaches. Pizza has known and achieved this for years. Control your destiny, control your delivery. By vertically integrating order, production, and fulfillment, your restaurant can prevent the problems that come with third-party involvement.
About the Authors
Meredith Sandland and Carl Orsbourn are co-authors of “Delivering the Digital Restaurant: Your Roadmap to the Future of Food” and “Delivering the Digital Restaurant: The Path to Digital Maturity.” After each spent 20-plus years in corporate strategy and retail food, Meredith and Carl concluded that food in America was changing. They left their corporate jobs in search of innovation that would transform the restaurant industry. 
Ghost kitchens, virtual brands, digital marketing, the gig economy and lean operations are at the heart of the future they envision. Meredith is the CEO of Empower Delivery, software that powers delivery-centric kitchens. Carl is the co-founder of Juicer and an advisor to restaurant groups and technology solutions. Subscribe to their newsletter and podcast at deliveringthedigitalrestaurant.com.