DoorDash: How Paper Routes Go Digital
The Paper Route
Tony Xu spent a week delivering pizzas for Domino’s not as performance art for investors, but as due diligence. He wanted to understand why restaurant delivery had historically been a graveyard of capital. What he found was an operation run on paper, printed tickets, phone calls, static routes, managed by world-class logistics companies that happened to lack the smartphone-era coordination tools available to any teenager with Find My Friends. The insight was simple and economic: Domino’s and FedEx operated on fixed-cost labor models where drivers sat idle between orders, bleeding wages during the 2:00 PM to 5:00 PM dead zone. If you could instead aggregate demand across hundreds of small restaurants and route drivers dynamically using real-time location data, you could convert those fixed costs into variable ones. The labor would scale with demand rather than anticipate it.
This was 2013. Xu had spent the previous summer at Square, where he watched Jack Dorsey’s team use smartphone hardware to unlock card payments for merchants previously trapped in cash economies. The leap was obvious: apply the same logic to delivery logistics. Instead of requiring restaurants to hire fleets and manage dispatch, DoorDash would own the fulfillment layer, the Dasher network, and charge restaurants a commission for orders they could never otherwise fulfill. The initial model was crude but revealing: a $6 flat delivery fee to consumers, commissions from restaurants, and drivers paid via Square readers at the point of delivery, routed through iPhones running Find My Friends. The first order came within an hour of launch, placed by Bruce Barcott, a visiting author who wanted Thai food delivered to Stanford. The economics were untested at scale, but the mechanism was clear. DoorDash would make money only if it could keep drivers busy enough across the entire day to justify the gig-economy payment structure, while reaching restaurants too small or unsophisticated to build their own delivery operations.
The Utilization Engine
The difference between DoorDash and its early rival Grubhub was not marketing or app design; it was the location of the logistics burden. Grubhub operated as a demand-aggregation platform, taking a commission for routing orders to restaurants that handled their own delivery. This limited Grubhub to pizzerias and Chinese restaurants, stablishments that already employed drivers and owned the capital equipment necessary for fulfillment. DoorDash, by contrast, inserted itself into the logistics layer, hiring independent contractors to move food from kitchens that had never previously delivered. This expanded the total addressable market to include the long tail of small operators, your local Thai place, the boutique sandwich shop, the family-owned taqueria, but it also introduced a brutal operational constraint. The model only worked if drivers spent most of their time moving, not waiting.
The mathematics of delivery economics are unforgiving. A driver earning $15 per hour in guaranteed wages, working a six-hour lunch and dinner shift with two hours of idle time between peaks, requires a subsidy from the platform of $30 per shift just to break even on labor. If that same driver works as an independent contractor, paid only per delivery, the platform bears no cost during idle periods, ut the driver will abandon the platform unless they can achieve sufficient throughput to earn a living wage. DoorDash solved this by aggregating demand across dozens of restaurants in dense geographies, ensuring that a driver could complete a delivery for one restaurant, drive three minutes, and pick up an order from another. The smartphone enabled real-time coordination that made this density actionable: custom dispatch logic, Google Voice integration for masked phone numbers, and routing algorithms that treated drivers as dynamic assets rather than assigned territories.
This aggregation created the “utilization curve” that underpins DoorDash’s unit economics. By serving the lunch rush at the office complex, the 3:00 PM coffee run, and the dinner surge from suburban families, the platform could keep drivers occupied across non-peak hours that would bankrupt a restaurant employing dedicated staff. The variable-cost structure, paying per delivery rather than per hour, only achieves positive returns when utilization exceeds a threshold that requires geographic concentration. Without density, drivers spend twenty minutes driving between orders, earning below minimum wage and churning out of the platform. With density, the model approaches the efficiency of a utility, extracting commission revenue from restaurants that previously had zero delivery capability while maintaining labor costs that flex precisely with consumer demand.
The Geography Trap
Here lies the defining structural constraint: the very density required for economic viability conflicts with the geographic expansion necessary to achieve venture-scale returns. DoorDash raised $2.5 billion from SoftBank’s Vision Fund and sovereign wealth funds by December 2019 on the promise of national dominance. But high utilization requires clustering. Expand too fast into suburban or tertiary markets where order volume remains sparse, and driver utilization collapses. The Dasher network dissolves as earnings per hour drop below alternative gig work. Conversely, expand too slowly, and Uber Eats or Grubhub capture the market first, locking in restaurants and consumers through exclusive partnerships or brand recognition.
This tension shaped DoorDash’s pre-pandemic strategy. By late 2019, the company was tripling year-over-year revenue, aggressively buying market share in cities where it had not yet achieved the order density required for positive unit economics. The bet was that scale would precede profitability, hat losses in new markets were customer acquisition costs amortized over future density. But density is not merely a function of time; it depends on local population patterns, restaurant concentration, and commuting cultures. A market like Manhattan achieves viable utilization organically. A sprawled Sun Belt city may never achieve the order concentration necessary to justify the service, regardless of how long DoorDash operates there. The company is effectively running a portfolio bet: that enough markets will achieve critical density to subsidize the inevitable failure of others, and that the winners will generate sufficient commission revenue to justify the aggregate capital burn.
The Commission Squeeze
Between the driver seeking maximum earnings per hour and the consumer seeking minimum fees sits the restaurant, paying commissions that typically range from 15% to 30% of order value. For a restaurant operating on 5% net margins, a 30% commission on delivery orders creates a paradox: the incremental revenue is technically accretive to top-line growth, but potentially destructive to bottom-line profitability. DoorDash’s structural advantage, serving restaurants too small to manage their own delivery, simultaneously creates a ceiling on pricing power. These are precisely the operators most sensitive to margin compression.
The platform faces a three-sided marketplace tension that grows more acute as the company matures. Raise commissions to improve margins, and restaurants defect to Grubhub (where they handle their own delivery and pay lower fees) or attempt to build captive delivery systems using Shopify-style logistics tools. Lower commissions to retain restaurants, and the platform cannot sustain driver payments during peak hours without increasing consumer fees, which suppresses demand. The equilibrium requires that DoorDash maintain a spread, between what restaurants pay, what consumers pay, and what drivers earn, that satisfies all three parties sufficiently to prevent churn. This spread is narrow. In competitive markets, Uber Eats can undercut DoorDash’s commission rates temporarily, forcing a race to the bottom that destroys unit economics. The fragility is structural: DoorDash’s value proposition to restaurants rests on the premise that incremental delivery revenue justifies the commission, but if the commission itself erodes the restaurant’s profitability, the relationship becomes extractive rather than symbiotic.
Purchased Growth
The company claims to be the only unit-economics-positive standalone logistics player in food delivery. This claim rests on a fragile assumption: that current levels of consumer subsidy are temporary acquisition costs rather than permanent features of the business model. DashPass subscriptions, free delivery promotions, and first-user discounts have pulled forward demand, particularly during the pandemic years when in-person dining collapsed. The question is whether these subsidies have created genuine habit formation or merely rented consumer attention.
If DoorDash can remove subsidies without experiencing significant churn, the unit economics are genuinely positive. If removing subsidies causes order frequency to collapse, if consumers revert to cooking or pickup when faced with true delivery costs including service fees, delivery fees, and marked-up menu prices, then the claimed profitability is illusory. The business is simply purchasing growth with venture capital, deferring the moment of reckoning until public market pressure forces price rationalization. The pandemic created a tailwind that may have permanently altered consumer behavior, or it may have created a high-water mark for delivery demand. If demand reverts to 2019 levels as dining rooms remain open, DoorDash must justify its valuation on mature-market economics where growth has stalled but the infrastructure of driver networks and technology investment remains. The implicit bet is that the density achieved during the subsidy era creates network effects sticky enough to retain users at full price. The risk is that the density itself was an artifact of the subsidies, and that true demand at cost-covering prices is insufficient to maintain driver utilization.
The Technical Asymmetry
Most analyses of DoorDash’s competitive position focus on brand recognition or restaurant selection. The underappreciated structural advantage is technical and operational: the company built a smartphone-native logistics stack from inception, while Grubhub never needed to develop these capabilities. Grubhub’s demand-aggregation model required only a website and a phone line to restaurants that managed their own drivers. DoorDash’s fulfillment ownership required real-time coordination technology, Find My Friends for location tracking, Square integration for instant driver payments, custom dispatch algorithms, that restaurants cannot replicate without significant engineering investment.
This creates an asymmetry that is difficult to reverse. A restaurant attempting to build in-house delivery must replicate not merely the app interface, but the routing logic, the supply-demand balancing algorithms, and the driver incentive structures that DoorDash has refined over a decade. Grubhub, attempting to pivot to full-service delivery, lacks the operational technology stack because its historical business model never required it. The density of DoorDash’s driver network is visible to competitors; the dispatch logic that maximizes utilization per driver is proprietary and cumulative. Every optimization, reducing wait times at specific restaurants, predicting prep times using kitchen data, batching orders to reduce per-delivery costs, widens the gap between DoorDash’s variable-cost structure and the fixed-cost alternatives. This is not a moat of brand loyalty, but a moat of operational complexity that appears simple to consumers but requires years of data and engineering to execute at margin.
The Employee Problem
The entire economic engine depends on gig labor remaining legally and economically viable. Proposition 22 in California provided temporary clarity, preserving independent contractor status for drivers while mandating limited benefits. But the legal status of gig workers remains precarious. If regulatory shifts reclassify Dashers as employees, the variable-cost advantage evaporates instantly. Employers must pay minimum wage during idle hours, payroll taxes, unemployment insurance, and workers’ compensation. The utilization problem that killed restaurant-owned delivery returns: drivers become fixed costs that bleed money during off-peak hours.
Even without legal reclassification, the economic assumption of infinite elastic labor supply may fail. During periods of low unemployment, or if competing platforms (Amazon Flex, Instacart, Uber) increase incentives, DoorDash must raise driver pay to maintain service levels. Since the company cannot easily raise restaurant commissions or consumer fees without triggering the defection dynamics described earlier, increased labor costs compress the narrow spread that generates platform profit. The model assumes that smartphone-enabled coordination allows DoorDash to pay less per delivery than a traditional employer would pay per hour, while still offering drivers sufficient earnings to participate. If this equilibrium shifts, if drivers demand higher compensation or if regulatory requirements increase costs, the mathematics of density no longer resolve into positive unit economics.
Operations as Margin
Tony Xu’s week at Domino’s was not merely founder mythology. It represented a cultural orientation that functions as a direct economic variable: the treatment of logistics as a technology problem rather than a labor management problem. This distinction shapes capital allocation. DoorDash invests heavily in dispatch algorithms, prep time prediction models, and batching logic that increase driver utilization by single-digit percentages. In a business where net margins might eventually reach 3-5%, a 2% improvement in utilization through better routing is the difference between profitability and perpetual loss.
The operational intensity extends to restaurant onboarding and quality control. By standardizing tablet interfaces, photography requirements, and menu formatting, DoorDash reduces the friction that slows order throughput. Every minute a driver waits at the curb is a minute of utilization lost. The company’s culture of measuring delivery times to the second, of penalizing restaurants with consistently long prep times, and of optimizing the handoff geometry between kitchen and driver, hese are not soft cultural factors but hard economic constraints. In the three-sided marketplace, operational efficiency is not merely a competitive advantage but a survival mechanism. The platform that moves food most efficiently extracts the maximum possible spread between restaurant commission and driver cost, while the inefficient platform sees its drivers churn and its restaurants defect.
DoorDash’s future depends on whether this operational rigor can sustain the density required to make gig labor viable, while navigating the geographic expansion necessary to satisfy capital markets: tension that grows more acute as the subsidy era ends and the true cost of delivery meets the true willingness of consumers and restaurants to pay.
The Cold Start Problem in Reverse
The geographic expansion dilemma reveals itself most starkly in markets like Phoenix or Dallas, ities with the population density to suggest viability but the automotive infrastructure to undermine it. In Manhattan, a Dasher might complete three deliveries per hour, moving from a Midtown salad concept to a Downtown poke shop to a West Village Italian spot, each trip measured in blocks rather than miles. In Phoenix, the same driver covers twelve miles between pickups, burning gas and time across highway segments that pay nothing. The model assumes that density creates utilization, but density is not merely a function of population; it is a function of urban form, parking availability, and the spatial distribution of restaurants relative to housing.
When DoorDash enters such markets, it faces the cold start problem in reverse. In the classic network effects model, a platform struggles to achieve initial liquidity, nough drivers to attract consumers, enough consumers to retain drivers. DoorDash solved this in its early markets through subsidy-funded guarantees: promising drivers minimum earnings regardless of order volume to ensure supply during the demand build phase. But in secondary markets, the problem persists chronically. The platform can buy driver presence, but it cannot buy proximity. If the Thai restaurant is in Scottsdale and the customer is in Tempe, the delivery takes forty minutes. The consumer, accustomed to the fifteen-minute standards of dense coastal cities, churns. The driver, having completed one delivery in an hour rather than three, abandons the platform for Instacart or Amazon Flex.
This creates a portfolio of impaired assets. By 2019, DoorDash operated in thousands of cities, but the unit economics varied by an order of magnitude between its top quartile and bottom quartile markets. The venture capital imperative to demonstrate total addressable market expansion encouraged land-grab strategies that masked this heterogeneity. The company reported aggregate growth metrics, tripling revenue, expanding market count, that obscured the divergence between the profitable Manhattan operations and the cash-incinerating exurban sprawls. The tension is existential: the markets that most need delivery logistics (car-dependent suburbs where cooking is inconvenient but driving is universal) are precisely those where the utilization math fails. The markets where utilization works (dense urban cores with parking scarcity) are increasingly saturated with competition and regulatory scrutiny.
The Restaurant Revolt
The commission structure that funds this expansion creates its own feedback loop of resistance. At fifteen percent, a restaurant treats delivery as incremental margin; at thirty percent, it treats the platform as a necessary evil to be minimized. The response has been gradual but structural. In the early years of the delivery boom, restaurants accepted high commissions as the cost of accessing demand they could not generate independently. As the platforms matured, restaurants recognized that DoorDash had inserted itself as a toll collector between the merchant and their own customers.
The countermeasures are technical and organizational. Toast, the restaurant payment processor, and Shopify, the e-commerce infrastructure provider, have built native delivery logistics modules that allow restaurants to offer direct ordering with third-party driver dispatch: ffectively decoupling the demand generation from the fulfillment. A restaurant can maintain its own customer relationship, capture the email and phone data, and still access gig-economy drivers through APIs that bypass the DoorDash marketplace. This vertical disintegration threatens the commission revenue that subsidizes DoorDash’s technology investments. If restaurants capture the demand organically through Instagram or Google Business, and fulfill through white-label logistics providers, DoorDash becomes a commodity router of labor rather than a value-creating marketplace.
The resistance has also taken political form. During the pandemic, cities including San Francisco, New York, and Seattle imposed temporary caps on delivery commissions, typically limiting them to fifteen percent. These interventions revealed the fragility of the model’s margin structure. DoorDash responded by adding consumer fees explicitly labeled “regulatory response” or by reducing service quality in capped markets, longer wait times, fewer drivers, to restore the economic spread by degrading the product rather than absorbing the cost. The political dynamic suggests that restaurants, when organized, possess countervailing power that the platform cannot permanently suppress. The thirty-percent commission that funded growth may be politically unsustainable in major markets, while economically unsustainable at lower rates.
The Peak Hour Crunch
The driver supply problem manifests most acutely in the asymmetry between lunch and dinner demand. Dinner represents a extended surge, amilies ordering between 6:00 PM and 9:00 PM, with enough volume to keep drivers occupied continuously. Lunch is compressed: office workers ordering in a narrow 11:30 AM to 1:30 PM window, with precipitous demand collapse before and after. The platform must maintain sufficient driver supply to serve the peak without overpaying for the trough, but gig workers optimize their own earnings per hour, not the platform’s service levels.
During peak lunch hours, DoorDash faces an elasticity problem with no clean solution. If surge pricing is applied to attract drivers, the consumer fees rise to levels that suppress order volume, welve-dollar delivery fees for a fifteen-dollar salad convert the transaction from convenience to luxury, and demand drops precipitously. If surge pricing is withheld, delivery times extend to forty-five minutes, and the office worker walks to the corner deli instead. The driver, observing the long waits and short delivery windows, switches to Uber Eats or Grubhub where the immediate incentive structure appears more favorable.
The multi-apping phenomenon exacerbates this. Experienced Dashers run four applications simultaneously, DoorDash, Uber Eats, Instacart, Amazon Flex, accepting the highest-paying offer in real-time and pausing the others. This labor arbitrage prevents DoorDash from building captive supply. A driver may appear “active” on the DoorDash platform but decline six consecutive orders because better compensation appears on competing apps. The platform sees utilization in the aggregate data, total active drivers versus total orders, but misses the microstructure of driver behavior: the cherry-picking of high-value orders, the rejection of low-margin trips to distant suburbs, the abandonment of the platform during low-incentive periods.
The operational response, algorithmic management of driver incentives, heat maps showing “busy” zones, quest promotions for completing a certain number of deliveries, creates a cat-and-mouse game that increases the effective labor cost without securing loyalty. The platform must pay more to maintain service levels during peaks, but cannot raise consumer prices sufficiently to cover these costs without destroying demand. The result is margin compression during the very hours when the business should be most profitable.
The Capital Intensity Trap
The gig-economy model was sold to investors as capital-light compared to traditional logistics. FedEx owns trucks and pays benefits; DoorDash owns nothing and pays per delivery. But the reality has proven more complex. The fulfillment ownership that differentiates DoorDash from Grubhub requires continuous investment in routing technology, mapping data, customer service infrastructure, and fraud prevention. When a driver delivers the wrong order or a restaurant closes unexpectedly, DoorDash bears the cost of remediation, efunds, credits, redelivery. The variable-cost labor model does not eliminate fixed-cost infrastructure; it merely shifts the fixed costs from vehicles and warehouses to software and operations centers.
Compare the capital structure to Domino’s, the company Xu studied so carefully. Domino’s owns its stores, or franchises them with strict operational controls, and maintains a fleet of company-owned delivery vehicles in many markets. This appears capital-intensive, but the assets generate returns for decades. DoorDash’s technology investments depreciate faster; the routing algorithms of 2018 are obsolete by 2024, replaced by machine learning models that require new data infrastructure and engineering talent. The company employs thousands of software engineers, data scientists, and operations managers whose salaries constitute fixed costs that do not flex with delivery volume. The “capital-light” narrative obscures the reality that DoorDash has built a technology company with logistics margins: mismatch between the valuation multiples of software and the economic returns of delivery.
The infrastructure burden extends to physical assets as well. DoorDash has experimented with ghost kitchens, enting warehouse space where multiple virtual restaurants share cooking facilities and Dasher pickup locations. This represents a retreat from the asset-light model toward the very fixed-cost real estate that characterized failed delivery ventures of the 1990s. The logic is defensive: by controlling the pickup location, DoorDash reduces wait times and increases throughput. But the economics revert to traditional real estate: lease obligations, utility costs, and the operational complexity of managing food preparation rather than merely routing it.
The Demand Elasticity Question
The bear case rests ultimately on consumer price sensitivity. The delivery industry has operated for a decade under the assumption that convenience commands a permanent premium, hat consumers will pay eight to twelve dollars in fees (delivery, service, tip) on top of marked-up menu prices to avoid leaving their homes. This assumption has not been tested at scale during normal economic conditions. The pandemic years created an artificial environment where the alternative to delivery was not cooking or dining out, but eating pantry staples in isolation. As dining rooms reopened, the true price elasticity of delivery demand began to reveal itself.
In economic downturns, the convenience tax faces its severest test. When household budgets contract, the marginal decision between cooking pasta and ordering Thai delivery shifts dramatically. The delivery fees that seemed trivial during prosperous months become prohibitive when inflation erodes discretionary income. DoorDash’s response, DashPass subscriptions that reduce per-order fees in exchange for monthly payments, attempts to lock in commitment through sunk-cost psychology. But the subscription model merely defers the reckoning. If the subscriber orders twice per month rather than six times, the effective per-order cost still rises, and the rational consumer cancels.
The platform’s historical losses, funded by venture capital, created an illusion of sustainable demand. Consumers experienced artificially low prices, ubsidized by investors seeking market share rather than profit. As the company approaches breakeven, it must raise prices or reduce subsidies to levels that may not sustain the order volumes necessary for driver utilization. The density that makes the model work depends on frequency: the same customer ordering three times per week rather than once. If price rationalization reduces frequency to once per week, the driver network loses the order density required to maintain earnings, triggering the utilization death spiral.
The ultimate question is whether DoorDash has built a genuine logistics utility or merely a temporarily subsidized convenience. If the former, the company will survive the withdrawal of subsidies, achieving profitability through operational excellence and scale economies that competitors cannot match. If the latter, the business is a transfer of wealth from investors to consumers and drivers, temporarily masking an economic reality where food delivery costs more than consumers are willing to pay and restaurants are willing to bear. The coming years will determine whether the smartphone-enabled routing that Xu identified at Domino’s represents a permanent efficiency gain or a momentary arbitrage between venture capital abundance and labor regulatory ambiguity. The density required to make the math work may exist only in the presence of subsidies that the underlying economics cannot sustain.
The comparison with Domino’s reveals the depth of the economic mirage. Domino’s operates on a captive demand model: consumers order through proprietary channels, the company owns the entire production and delivery stack, and the labor force is employed rather than gigged. This appears inefficient on a spreadsheet, fixed wages, vehicle maintenance, real estate costs, yet it generates sustainable margins because the demand is organic and the pricing reflects true cost. A Domino’s delivery in suburban Ohio costs the consumer less than a DoorDash delivery from the same pizzeria because the vertical integration eliminates the platform spread. DoorDash’s intermediary position adds a structural cost that can only be justified by incremental demand, rders that Domino’s cannot fulfill because they lack a Thai kitchen or sushi bar. But if the incremental demand requires thirty percent commissions and ten-dollar delivery fees to sustain, it is not clear that the demand actually exists in a form that supports a standalone business.
The ghost kitchen experiments represent an implicit acknowledgment of this constraint. By leasing industrial warehouse space and subdividing it into kitchen stalls for virtual brands, restaurants that exist only on DoorDash’s app, the company attempts to arbitrage its own commission structure. If DoorDash cannot extract sufficient margin from third-party restaurants without triggering defection, it must capture the margin itself by becoming the restaurant. This reverses the original capital-light thesis. Suddenly the company holds ten-year leases on speculative real estate in exurban logistics parks, installs grease traps and ventilation systems, and manages health inspections for entities with no customer-facing presence. The economics revert to those of a franchise operator, with the added complexity of coordinating delivery logistics for multiple competing concepts sharing a single loading dock. The density advantage, the very characteristic that made the model seem clever, becomes a liability when the platform must fill kitchen capacity across the same off-peak hours that bedevil driver utilization. A ghost kitchen serving lunch bowls at noon and pizza at midnight requires the same fixed-cost coverage as a traditional restaurant, but without the brand recognition or dine-in revenue to amortize the rent.
The financial reporting obscures these tensions through the architecture of “adjusted” metrics. DoorDash reports “contribution profit” at the marketplace level, revenue minus direct costs, which excludes the corporate overhead, technology development, and marketing spend required to generate that revenue. By this measure, an individual delivery in Manhattan appears profitable while the same delivery in Phoenix appears merely unprofitable rather than catastrophic. But the corporate costs are not optional; they are the price of maintaining the routing algorithms and driver incentives that make the marketplace function. When analysts model the path to true net income, they must assume that corporate overhead grows slower than revenue, that marketing spend declines as brand awareness saturates, and that technology investment reaches steady-state maintenance rather than continuous escalation. Each assumption contradicts the competitive dynamics of the industry. As Uber Eats and Instacart improve their routing capabilities, DoorDash must accelerate its technology investment merely to maintain parity, not to achieve dominance. The marketing spend required to acquire a new user in a mature delivery market, where every consumer has already tried the major platforms, exceeds the lifetime value of that user if subsidies are removed.
The pandemic created a specific distortion that temporarily resolved these contradictions. When restaurants closed for dine-in, they accepted any commission rate because thirty percent of a delivery order exceeded zero percent of a closed dining room. Consumers accepted inflated fees because the alternative was cooking at home during lockdown. Drivers accepted lower effective wages because alternative employment had evaporated. This truce, where all three sides of the marketplace accepted suboptimal terms due to external shock, generated the metrics that justified the IPO valuation. But the truce was temporary. As restaurants reopened, they began negotiating commission reductions or building direct ordering channels. As employment options expanded, drivers multi-app’d more aggressively, forcing the platform to increase incentives. As consumers regained mobility, they reduced order frequency, concentrating remaining demand further into peak hours and exacerbating the utilization problem.
In this light, the “unit economics” that DoorDash claims represent not a sustainable equilibrium but a local maximum achieved during abnormal conditions. The density that permits positive contribution margin exists only when demand is artificially elevated by circumstance or subsidy. Remove the subsidy, and the density dissipates; remove the density, and the unit economics invert. The business is not a logistics network approaching scale efficiency, but a demand-coordination mechanism that requires continuous capital injection to prevent collapse. The smartphone-enabled routing that Tony Xu identified at Domino’s solved the technical problem of coordination without solving the economic problem of cost coverage. Until the platform can demonstrate profitability in a mid-sized market, say, Milwaukee or Nashville, without consumer subsidies or driver incentives above minimum wage, the model remains unproven. The current valuation rests on the assumption that the Manhattan density can be replicated everywhere, that the gig labor supply is infinite and frictionless, and that restaurants will indefinitely pay premiums for incremental demand that may not actually be incremental to their bottom line, but merely shifted from profitable dine-in to marginally profitable delivery. If any of these assumptions fail, the three-sided marketplace collapses into a conventional logistics business with conventional logistics margins, hat is, a business worth fractions of its current enterprise value.