Delivery driver bag in front of restaurant entrance
DeliveryOperationsMargins

In-house delivery vs DoorDash Drive

DoorDash Drive and Uber Direct fees vs running your own drivers — the real math, the ops cost, where FoodyOS dispatch fits, and when to keep a fleet.

FoodyOS Team
Operations
·10 min read

For most US restaurants, “delivery” means handing the order to DoorDash, Uber, or Grubhub. That’s changing — slowly — as more operators look at running their own driver fleet. The question isn’t whether self-managed delivery is cheaper per order (sometimes it is, often it isn’t once you account for time) — it’s about who owns the customer relationship. The first-order math behind that ownership question is in direct ordering vs DoorDash & Grubhub.

The numbers, plainly

Marketplace delivery commission is set by the consumer-facing platform. As published on each vendor’s own US pricing page in August 2026: DoorDash lists Marketplace tiers at 15%, 25%, and 30% of order subtotal,1 Uber Eats lists 20%, 25%, and 30% Marketplace fees plus a 15% self-delivery tier,2 and Grubhub publishes marketing commissions of 5%, 15% and 20%, with its own delivery fleet starting at another 10% on top.3 Bring-your-own-delivery (BYOD) products are priced separately: DoorDash publishes Drive On-Demand at $6.99 to $10.99 per delivery order, a flat fee with no commission rate,4 and Uber Direct is a per-delivery fee — Uber publishes it as starting at $7.99 per delivery — with no Marketplace commission, since the order originates on your channel.5 So if a $50 delivery comes in:

  • Marketplace order, marketplace driver: roughly $7.50–$15 in commission depending on the tier you signed for. You keep the rest minus food cost.
  • Direct order, marketplace driver via DoorDash Drive or Uber Direct: a flat per-delivery fee, no Marketplace commission. You keep the order subtotal minus that fee, your payment processing, and food cost.
  • Direct order, your own driver: driver wage plus mileage reimbursement. The actual cost depends on how you classify the driver and how many drops you batch — see the next section.

The win on owning your fleet isn’t huge per order. The win is at scale: you’re paying a marginal labor cost instead of a percentage that grows linearly with ticket size.

How drivers actually get paid

There are three ways a US restaurant ends up with a driver on the road:

  • W-2 employee. You pay hourly wages, payroll taxes (FICA, FUTA, state unemployment), and — in tip-credit states — at least the federal tipped minimum of $2.13 per hour with tips making up the rest of the standard $7.25 minimum.6 You also have to carry hired-and-non-owned-auto coverage on top of general liability, because a personal auto policy almost always excludes commercial delivery.7 The IRS standard business mileage rate, which most operators use to reimburse drivers tax-free, is published annually — for 2025 it is 70 cents per mile.8
  • 1099 independent contractor.Cheaper on paper, much riskier on classification. The US Department of Labor’s 2024 final rule on employee vs. independent contractor status under the FLSA uses a six-factor economic reality test, and a driver who only delivers for one restaurant on a fixed schedule is hard to defend as truly independent.9 Signing a 1099 does not settle the question — the DOL test weighs the economics of the relationship, not the paperwork, and if it comes out the other way the back wages, payroll taxes and penalties land on the restaurant, not on the driver.
  • Gig fleet via BYOD (DoorDash Drive, Uber Direct, Grubhub Direct, Olo Dispatch).The driver is the platform’s contractor. You pay a per-delivery fee, the platform handles pay, insurance, and background checks. You give up GPS-level control and a thin slice of margin in exchange for zero fixed cost.

The four BYOD options, compared

  • DoorDash Drive.DoorDash’s courier network sold on its own, separate from a Marketplace listing — a published $6.99–$10.99 flat fee per delivery instead of a commission. Two flavors: Drive On-Demand (you push each order) and Drive API (you integrate dispatch into your stack).4 Coverage is decided ZIP by ZIP, so run your regular drop addresses through it before you commit.
  • Uber Direct. Same fleet that powers Uber Eats deliveries, exposed as a white-label API. Pricing is a per-delivery fee that Uber publishes as starting at $7.99.5
  • Grubhub Direct.Grubhub’s commission-free direct-ordering product, which can dispatch their own driver network as a delivery option.10 Useful if you already have Grubhub Marketplace and want to consolidate operationally.
  • Olo Dispatch.Aggregator layer — rather than one fleet, Olo brokers each order across a panel of delivery service providers; Olo’s own page reports 96% of its brands are covered by two or more DSPs.11Best fit for enterprise-scale brands that already use Olo’s ordering platform.

Batching is where the math actually changes

A W-2 driver paid $20/hour all-in (wage + mileage reimbursement at the IRS rate + payroll burden) doing 2 deliveries an hour costs $10 per order. The same driver doing 4 batched deliveries in the same hour costs $5 per order. Push that to 6 batched drops in an hour during a Friday pizza rush and your per-order labor cost falls to about $3.33 — below the bottom of DoorDash Drive’s published $6.99–$10.99 range4and below Uber Direct’s published $7.99 starting fee.5 The leverage is entirely in the second half of that fraction (deliveries), not the first (cost): the wage barely moves the number, and utilization moves it threefold.

Batching only works when drops are clustered, which is why pizza chains can run lean fleets and a single-location bistro with a 5-mile zone usually can’t. The practical test: look at your last 200 delivery tickets, plot the drop addresses, and ask whether any 4 of them could have left the kitchen in the same insulated bag without any of them arriving cold. If the answer is “rarely,” stop reading and just plug DoorDash Drive into your direct-ordering site. If the answer is “most Fridays,” you have a real argument for a fleet.

Two more line items to budget that nobody warns first-time operators about. First, the IRS standard mileage rate of 70 cents per mile in 2025 is the easy way to reimburse a personal vehicle without it becoming taxable wages,8but it has to be paid against a contemporaneous mileage log — “we’ll figure it out at year-end” is how you end up with a payroll-tax assessment. Second, hired-and-non-owned-auto coverage on a commercial policy typically runs a few hundred dollars per vehicle per year on top of your existing general liability, and skipping it is the single most common way an owner-operator finds out their personal carrier excludes commercial use the day after an at-fault collision.7

When self-managed delivery makes sense

  • You already have drivers(pizzerias, high-delivery cuisines). The infrastructure exists. FoodyOS dispatch + zones is just better software for what you’re already doing.
  • Your delivery zone is dense. If 80% of orders fall within 2 miles, batching becomes possible and a driver can do 4 deliveries in 30 minutes.
  • Your average ticket is high. Self-fulfilled delivery is a percentage win — the bigger the ticket, the larger the absolute dollar savings vs. marketplace commission.
  • You want the customer relationship.Marketplace delivery hides the customer’s phone number. Direct delivery means you can text them about a new menu next month.

When marketplaces are still the right call

  • You’re a single-location quick-service spot doing 10 deliveries a week. Don’t hire a driver. Use DoorDash Drive on direct orders.
  • Your delivery zone is sprawling and orders are rare. The driver utilization math kills you.
  • You’re launching and need volume. Use marketplaces for acquisition; convert to direct with bag stickers.

What a real self-delivery operation looks like in FoodyOS

The pieces, all on one platform:

  • Zones drawn on a map. Concentric rings around your restaurant, or hand-drawn polygons for tricky neighborhoods. Each zone has a delivery fee, a minimum order, and an ETA. Out-of-zone checkouts are politely rejected at the menu.
  • Drivers as users.Each driver gets a FoodyOS login. They run the driver app on their own phone — no proprietary handheld. Standard role: see assigned orders, hit Picked Up, hit Delivered. Optional: GPS-based status updates flow back to the customer’s tracking link.
  • Dispatch. Manual (the cashier picks who takes it) or automatic (the closest available driver). Drivers can batch — pick up 3 orders, drop them in optimal sequence.
  • Customer tracking link. The customer gets a link at order confirmation. They reopen it whenever they want to see where the driver is. No SMS spam, no marketplace branding.
  • Per-driver stats.Orders today, average dispatch time, customer ratings (if you collect them), miles driven. Helpful for paying drivers fairly, helpful for figuring out who’s your fastest pizza driver.

Hybrid is usually the right answer

For most delivery restaurants the answer is not one or the other. Your own drivers cover the high-volume zones during peak hours, where density makes an hourly driver cheaper per drop than a per-order fee; a white-label fleet service like DoorDash Drive or Uber Direct — which you contract directly, without listing on their marketplace — absorbs the spikes and the rare order to a far zone. Your fleet takes the dense middle of the curve, the contracted fleet takes the long tail. FoodyOS runs the in-house side of that: direct orders, your drivers, your dispatch queue, no commission. Handing an order off to a third-party fleet from inside FoodyOS is not integrated today — for now that stays in the provider’s own portal, and the two are managed side by side.

The wrong answer is the default — letting all your delivery flow through marketplaces. They get the customer, they get the margin, they keep the brand. Even if you keep their drivers, they don’t need to keep their commission. Direct dispatch is included in the FoodyOS platform — see pricing, or zoom out to FoodyOS vs Grubhub for the commission contrast at the platform level. The companion piece direct ordering vs DoorDash & Grubhub walks through the consumer-side commission math in more depth.

Sources

  1. DoorDash for Merchants, “Merchant Fees & Pricing” — Basic 15%, Plus 25%, Premier 30% on delivery; 6% pickup on all plans. Accessed August 2026. merchants.doordash.com/en-us/pricing.
  2. Uber Eats, “Pricing That Works for Your Business” — Marketplace fees 20% / 25% / 30%, self-delivery 15%. Accessed August 2026. merchants.ubereats.com/us/en/pricing.
  3. Grubhub for Restaurants, “Grubhub pricing and fees” — marketing commissions 5% / 15% / 20%, Grubhub Delivery “delivery fees start at 10%.” Accessed August 2026. get.grubhub.com/grubhub-pricing-and-fees.
  4. DoorDash, “Drive On-Demand” — “Drive On-Demand Pricing: $6.99 to $10.99 per delivery order … Flat delivery fee per order, no commission rates.” Accessed August 2026. merchants.doordash.com/en-us/products/drive.
  5. Uber Eats, “Pricing That Works for Your Business” — Uber Direct: “Starts at $7.99 per delivery.” Accessed August 2026. merchants.ubereats.com/us/en/pricing.
  6. U.S. Department of Labor, Wage and Hour Division, “Fact Sheet #15: Tipped Employees Under the FLSA,” dol.gov/agencies/whd/fact-sheets/15-tipped-employees-flsa.
  7. Insurance Information Institute, “Hired and Non-Owned Auto Insurance,” iii.org/article/business-auto-insurance-basics.
  8. Internal Revenue Service, “IRS issues standard mileage rates for 2025,” IR-2024-312, irs.gov/newsroom/irs-issues-standard-mileage-rates-for-2025.
  9. U.S. Department of Labor, “Final Rule: Employee or Independent Contractor Classification Under the FLSA,” 29 CFR Parts 780, 788, and 795, dol.gov/agencies/whd/flsa/misclassification/rulemaking.
  10. Grubhub, “Grubhub Direct — Commission-Free Online Ordering,” get.grubhub.com/products/grubhub-direct.
  11. Olo, “Restaurant Delivery Management Software | Olo Dispatch.” Accessed August 2026. olo.com/dispatch.
Run a high-volume US restaurant?
We'll show you FoodyOS in 30 minutes.
Book a demo
Photo: Unsplash