Latest / Restaurant Business with Fexingo: Food Service, Hospitality, and Independent Operators / Third-Party Delivery's Hidden Cost for Restaurants
Transcript
- Lucas: Let me start with a number that still bothers me. A $20 salad ordered through a delivery app — the restaurant gets about $9.50 after all the fees. That's before the cost of the ingredients, the labor, the rent. Luna: Nine fifty out of twenty? That's brutal. I knew commissions were high, but that seems worse than the standard thirty percent narrative. Lucas: It is. The thirty percent is just the starting point. You've got a commission fee — typically fifteen to thirty percent of the subtotal. Then a per-order fee, usually two to five dollars. Then a marketing fee if the operator wants visibility on the app, which can be another five to ten percent. And if the customer doesn't tip enough, the restaurant sometimes covers part of the driver's pay through a 'dispatch fee.' Luna: So the headline commission is like the cover charge. The real cost is the cumulative stack. Have you seen a breakdown for a specific order that makes this concrete? Lucas: Yeah, I pulled data from a 2025 industry report by the National Restaurant Association. They looked at a typical pizzeria order: a $25 large pizza with two toppings. On DoorDash, the commission was 25 percent of the subtotal — six twenty-five. Then a $2.99 per-order fee. Then a $1.50 marketing fee. That's ten seventy-four in fees. The restaurant's net is fourteen twenty-six. But the pizza itself costs about six dollars in ingredients and maybe four in labor. So the operator makes four dollars and change on a transaction that the customer thinks cost them thirty bucks with tip and delivery fee. Luna: So the operator is effectively working for a four-dollar margin. That's not sustainable for most independents. Does the math change for fine-casual concepts with higher ticket prices? Lucas: It improves slightly, but not enough. A $45 steak entree at a fine-casual spot — commission at thirty percent is thirteen fifty, per-order fee three fifty, marketing fee two dollars. That's nineteen dollars in fees. Net is twenty-six. Food cost might be twelve, labor eight. So profit is six dollars. On a forty-five-dollar sale. The operator is making thirteen percent margin on the delivery channel while the dine-in channel might net twenty-five to thirty percent. Luna: There's this term I've seen — 'delivery margin inversion.' Is that what we're describing? The delivery channel earns a lower margin than dine-in, forcing the operator to raise dine-in prices to subsidize the delivery customer? Lucas: Exactly. That's the inversion. And it's why you see restaurants charging twenty-two dollars for a burger on DoorDash that costs sixteen at the counter. The delivery menu prices are higher to absorb the fees, but that creates another problem — customers feel gouged. A survey last year found that sixty-eight percent of delivery customers think the menu prices on apps are 'unfairly higher' than dine-in. So they tip less or don't reorder. Luna: Let me push back a little. The platforms argue they provide marketing, order processing, and delivery logistics. Some operators say they wouldn't have the volume without them. Is there any scenario where the math works for the restaurant? Lucas: There is, but it's narrow. If the restaurant has very high average order values — say, a family meal for four at seventy-plus dollars — the fixed per-order fees become a smaller percentage. Also, if the operator uses the platform purely for incremental volume during off-peak hours when the kitchen is already staffed and underutilized, the marginal cost of that order is lower. But for most operators, especially independents doing under a million a year in revenue, delivery is a loss leader they can't afford to drop because competitors are on the apps. Luna: So they're trapped. That's the classic platform dynamic. What about the alternatives we're starting to see? I've heard about cooperative delivery networks — restaurants teaming up to share a delivery fleet. Does that pencil out? Lucas: It can, but the scale is still small. There's a group in Portland called the Independent Delivery Alliance — about forty restaurants share a pool of five drivers. They use a simple ordering platform that costs fifty dollars a month plus a flat two-dollar fee per order. The restaurant sets its own delivery radius. The driver gets a guaranteed wage plus tips. The math works if the order density is high enough — meaning multiple orders per hour per driver within a compact geographic area. Luna: How does the customer experience compare? I'd imagine the app isn't as polished as DoorDash or Uber Eats. And the selection is limited to those forty restaurants. Lucas: You're right. The UX is clunkier — custom orders require a phone call sometimes. But the restaurants report that customers who use it are more loyal and tip better. One pizzeria owner in the alliance said his net per order on the cooperative channel is eighteen dollars versus nine on DoorDash. That difference keeps him alive. Luna: Eighteen versus nine — that's a hundred percent improvement. But can it scale beyond a niche? I think about the convenience expectation customers now have: they want to see every restaurant, compare, and order in thirty seconds. Lucas: That's the tension. The big platforms have conditioned customers to expect a vast selection and a seamless experience. A cooperative can't replicate that without massive investment. But there's a middle path: direct-ordering integrations. Some POS systems now offer a 'direct order' button on a restaurant's website or Instagram page that routes to the restaurant's own delivery staff or a shared fleet. No third-party marketplace. The customer gets convenience; the restaurant keeps the full margin minus a small processing fee. Luna: Does the customer adoption exist? I'm skeptical that most people will go to a restaurant's site instead of opening DoorDash. Lucas: Adoption is low — maybe five to ten percent of delivery orders currently go direct. But it's growing. Chains like Sweetgreen and Chipotle have built their own ordering systems and seen profitability on delivery improve by twelve to fifteen percentage points. For independents, the key is making the direct channel easy to find — putting a QR code on the takeout bag, offering a loyalty discount for direct orders, even a simple text-message ordering system. Luna: So the endgame might be a hybrid: use the big platforms for discovery and new customer acquisition, then nudge those customers to direct orders for repeat business. Is that what the smart operators are doing? Lucas: That's exactly the playbook I'm seeing from the more digitally savvy independents. They accept that the first order might come through DoorDash at a loss. But they insert a card in the bag saying 'Order direct next time and get ten percent off.' They capture the email or phone number. After three direct orders, they've recouped the acquisition cost. The lifetime value of a direct customer is much higher because the margin is better. Luna: Okay, but let's be realistic. How many mom and pop restaurants have the time or digital skills to run that playbook? Most owners I know are working twelve-hour days just keeping the line moving. Lucas: Valid point. That's where third-party services that specialize in 'direct order enablement' come in. Companies like Olo and Chowly offer white-label ordering pages that integrate with the restaurant's POS and handle the logistics for a flat monthly fee plus a small per-transaction fee — much lower than marketplace commissions. The restaurant gets a branded ordering page, automated driver dispatch if they have their own fleet, and customer data. It's a middle ground. Luna: So the solution isn't abandoning the platforms — it's using them more strategically. I want to circle back to the regulatory angle. We've seen cities like New York, San Francisco, and Seattle cap delivery commissions at fifteen to twenty percent during the pandemic, and some have made those caps permanent. Has that helped? Lucas: It has, but only partially. Commission caps do reduce the fee percentage, but platforms respond by adding other fees. In New York, after the permanent fifteen percent cap on delivery commissions, DoorDash introduced a 'regulatory response fee' of two dollars per order and increased the marketing fee. So the total cost to the restaurant only dropped from about thirty-two percent to twenty-five percent. Still an improvement, but not the full relief intended. Luna: Plus the platforms have pushed back legally, arguing the caps violate their rights. And they've raised consumer fees to compensate. So the diner pays more. Is there any path to a sustainable equilibrium? Lucas: I think the equilibrium will be forced by margin pressure. Right now, the platforms themselves are barely profitable or losing money. DoorDash only turned its first net profit in 2024, and it was thin. Uber Eats is still unprofitable in many markets. Eventually, they need to raise prices to investors or find efficiencies. One possibility is that they move to a subscription model for restaurants — pay a flat monthly fee for lower per-order commissions. Some platforms are testing that with larger chains. Luna: Would independents benefit from that? A flat fee could be easier to budget for, but it might also lock them into a high cost if volume is low. Lucas: It depends on volume. For a high-volume pizza place doing two hundred delivery orders a week, a flat fee of, say, five hundred dollars a month plus a five percent commission could be cheaper than the current model. For a low-volume cafe doing twenty orders a week, a flat fee would be a disaster. So the platforms would likely segment pricing. Luna: Let me shift gears. I've been talking to a few operators who are experimenting with something called 'ghost kitchen partnerships' to make delivery profitable. They cook out of a shared commissary kitchen with lower rent, and the delivery is handled exclusively by the platform. Does that change the math? Lucas: It changes the fixed-cost side. Rent is lower, sometimes zero if it's a platform-owned ghost kitchen. But the variable costs — commissions — are the same or higher because the platform may charge a premium for using its kitchen. Plus, you lose the dine-in margin entirely. So it only works if you can drive very high order volume to offset the low per-order profit. Most ghost kitchen operators fail within a year because they can't get enough repeat orders. Luna: So it's not a panacea. What about the technology side? Are there any startups trying to solve the commission problem directly with blockchain or smart contracts? Lucas: There are a few, but they're very early. One project called BiteSwap uses a decentralized network where customers pay in crypto, and the smart contract automatically splits the payment between restaurant and driver with a one percent fee. It's interesting in theory, but the UX is terrible for the average diner. And volatility in crypto makes pricing difficult. I don't see it making a dent in the next few years. Luna: So we're stuck with a broken system for the foreseeable future. Let me ask you this: if you were advising a restaurant owner today, what's the single most actionable step to improve delivery profitability? Lucas: I'd say two things, but they're linked. First, raise delivery menu prices by at least fifteen percent compared to dine-in prices. Most owners are afraid to do this, but customers have been conditioned to expect it. Second, capture the customer data from every delivery order — phone, email, whatever — and start a simple loyalty program for direct orders. Even a punch card works. If you can convert twenty percent of your delivery customers to direct orders within three months, your overall delivery margin can improve by five to eight percentage points. Luna: That's surprisingly concrete. I've also heard that some restaurants are using dynamic pricing on delivery — charging more during peak hours — to offset the higher platform fees. Have you seen that work? Lucas: It's risky. Customers get angry when they see a surge on their favorite pasta dish. But some operators have done it subtly — a 'peak time surcharge' of two dollars on orders placed between seven and nine PM. The key is transparency. If you explain that the surcharge goes to paying drivers a fair wage during rush, customers are more accepting. We did an episode on dynamic pricing a few weeks ago, and the data showed that disclosure reduces backlash by forty percent. Luna: Right, I remember that. So restaurants have tools, but they require guts and communication. Final thought: where do you see the delivery landscape in three years, say 2029? Lucas: I think we'll see a bifurcation. The big platforms will continue to dominate for on-demand, broad-selection ordering, but they'll be forced to unbundle their fees and offer more transparent, tiered pricing for restaurants. Meanwhile, a patchwork of local cooperatives and direct-ordering integrations will grow, especially in dense urban areas. The margin will never be as good as dine-in, but it may become sustainable for more operators. The ones who survive will be those who treat delivery as a distinct channel with its own pricing, marketing, and operations — not just an afterthought. Luna: That sounds like the same playbook every industry has learned from e-commerce: own your customer. Restaurants are just late to the party. Thanks, Lucas. Lucas: Always a good conversation. Next episode we're looking at the resurgence of the neighborhood bakery — a completely different margin story.