What Third-Party Delivery Actually Costs You (The Math Nobody Quotes)

When DoorDash signed up our restaurant in 2021, the rep walked me through a slide deck that quoted a 15% commission. Sounded fine. The math on a $30 ticket said we’d net about $25.50, so we signed. Six months later I pulled the actual deposit data against the actual ticket count, and the average net per order was $19.40, not $25.50.

The 15% was real. So was the 35%. They were just describing different parts of the same order. Here’s the math you need before you sign anything.

The headline commission is not the all-in cost #

Every third-party platform leads with a commission rate. DoorDash quotes a tiered structure (15% Basic, 25% Plus, 30% Premier). Uber Eats sits in roughly the same range. GrubHub runs higher on the all-in number once delivery and marketing fees stack on (GrubHub published rates).

The commission is the headline. It’s not the all-in.

The all-in cost of a third-party order is the commission plus payment processing, the marketing tier you opted into to stay in the rotation, the in-app promotion you ran to compete with the restaurant next door, and the order-error refund rate. Add it all up and the typical all-in cost on a DoorDash Basic order runs 30-32% of order value. On Plus, 35-38%. On Premier with promotions running, 40-45% (Toast 2024 delivery economics study).

The restaurant net margin in the same year ran 3-9% (NRA 2024). The arithmetic only works one way. A delivery-only ticket frequently loses money.

Where the hidden fees come from #

There are six layers between the customer ordering and the deposit hitting your bank.

1. Headline commission. 15-30% depending on tier. The number on the contract. The number the rep quotes you.

2. Payment processing. Some platforms bundle this into the commission. Others bill it separately. Usually 2.9-3.5% of order total. On Uber Eats it is bundled. On DoorDash older contracts it was separate; on newer Marketplace contracts it bundles into the tier (DoorDash Merchant Terms). Read your specific contract.

3. Marketing tier opt-in. The “Plus” or “Premier” tiers buy you better placement in the app feed and access to DashPass / Uber One members. The fee gets bolted on as an additional commission point (5-10%). The conversion lift is real, usually 20-40% more orders, but the math is per-order: you pay 5-10% more on every order to get 20-40% more orders. Whether that pencils depends on your unit margin.

4. In-app promotion. Restaurant-funded promos (“20% off your first order,” “$5 off orders over $20,” BOGO items) come out of your pocket, not the platform’s. The platform funds the occasional sitewide promo. Your specific promo, almost never. Promotion spend can easily add another 5-10% on top of commission and processing.

5. Order-error refund rate. When a customer reports a missing item or a cold burger, the platform refunds them and bills you. Industry average runs 4-6% of orders flagging an issue, with average refund value around 25-40% of the ticket (Restaurant Business Online). Net effect on your gross: roughly 1-2% off the top.

6. Account-level fees. Setup fees, tablet rental, photography fees if you didn’t bring your own. Spread across a year, usually 0.5-1% of revenue.

Sum it: 15% commission + 3% processing + 7% tier + 6% promotion + 1.5% refund losses + 0.5% account fees = 33% all-in on a Plus-tier DoorDash account running modest promotions. That is the number that has to clear your food cost, your labor cost, your packaging cost, your kitchen overhead, and still leave a profit.

A worked example: where the $30 ticket actually lands #

Customer orders a $20 burger and $3 fries on DoorDash Plus. Adds a $5.99 platform delivery fee, a $2.50 service fee, and tips the driver $5.

The customer pays $36.49 total.

Of that, you, the restaurant, see only the food subtotal: $23.

The platform takes:

• Commission (25% Plus tier on $23): $5.75
• Payment processing (bundled, included in tier): $0
• Promotion the restaurant funded (“$3 off orders over $20”): $3.00
• Order-error rate amortized at 2%: $0.46

Restaurant’s net deposit on the order: $23 - $5.75 - $3 - $0.46 = $13.79.

The customer paid $36.49. You netted $13.79. That is 37.8% of what the customer paid.

Now run the unit economics on the $13.79:

• Food cost on $23 ticket at 30%: $6.90 COGS.
• Packaging (clamshell, fries box, bag, condiments): $1.10.
• Labor allocated per ticket (back-of-house portion, no front-of-house tip): $4.20.

Total ticket cost: $12.20. Net profit on this order: $1.59, or about 5% margin on the food subtotal you saw, but only 4.4% margin on what the customer actually paid.

Order that same $20 burger, $3 fries, and $5 tip in your dining room and the tip lands in a pool you don’t see, but the restaurant nets roughly $7 on the food after food cost and packaging-free service. The delivery order paid out one-fifth the profit of the dine-in version. Same burger.

The break-even check size for each platform #

The way to know if a platform makes money for you is the break-even ticket size: the order size where the platform fees plus your food cost plus your packaging plus your fulfillment labor exactly equals zero profit. Below that line you lose money. Above it you make money.

Rough break-even ticket sizes by platform (assuming 30% food cost, $1 packaging, $4 fulfillment labor, US-average commission tiers):

• DoorDash Basic (15%): $22-25 break-even
• DoorDash Plus (25%): $28-32 break-even
• DoorDash Premier (30%): $34-38 break-even
• Uber Eats standard (15%): $22-25
• Uber Eats Premium (30%): $34-38
• GrubHub all-in (~30%): $32-36

Your average ticket size decides whether a platform is a partner or a free ride for the platform. Average DoorDash ticket of $19 on Plus? Every order is a small loss. Average ticket of $44 and you’re making money even on Premier.

That’s why “lift your average ticket on third-party” beats “drop off third-party” for most operators. Minimum order thresholds, combo items, family meals, dessert add-ons, drink upsells. The gap between $19 and $32 is the gap between losing $1.50 a ticket and netting $4 a ticket.

When delivery actually makes sense #

There are four scenarios where third-party delivery pays.

1. You run with idle kitchen capacity. If your line is staffed for dine-in volume that doesn’t show up Monday-Wednesday at 2-5pm, delivery fills the gap at marginal cost, because the labor is paid either way. At marginal cost, even thin orders clear a profit.

2. Your average ticket sits above $30. The break-even thresholds clear on their own at higher tickets. Steakhouses, sushi, family-style Italian, BBQ tend to pencil. Burger and taco shops with $14 average tickets tend not to.

3. You use delivery as a marketing channel. Some operators treat third-party as customer-acquisition, eating near-zero margin on the delivery order to buy the customer’s eventual dine-in visit. This works if you have the data to track the conversion and a dine-in margin high enough to absorb the channel cost. It almost never works as well as the marketing rep tells you it will. I’ve sat through that pitch. The conversion they promise and the conversion you can actually prove are two different numbers.

4. You have a delivery-optimized menu. Items that travel well, package well, and price for travel are a different menu than your dining room. Concepts that take delivery seriously build a separate menu and price it differently. Off-platform sites like ChowNow let you build that menu with first-party delivery economics (10-15% all-in versus 30-40% on third-party).

What to do today #

Three quick numbers, fifteen minutes of work.

Number one. Pull last 90 days of third-party deposits and divide by ticket count. That is your real average ticket size per platform. Compare against the break-even table above. If you are below the threshold for your current tier, either tier down or work on ticket size before doing anything else.

Number two. Calculate true take-home per platform: total deposits ÷ total customer-paid order value. That tells you whether the headline commission matches the all-in reality. The gap between contract commission and actual take-home is your hidden cost.

Number three. Pull a sample of 20 orders per platform and back into the unit economics one ticket at a time. Find your real per-order profit, multiply by ticket count, and you’ve got the channel’s actual contribution to your P&L. If it’s negative, the platform is paying you to advertise for them, not the other way around.

The delivery profit calculator on this site runs the all-in math for any single order on any of the major platforms. Commission tier, packaging cost, food cost percentage, fulfillment labor, promotion spend. It gives you the net dollar profit per order and the contribution margin percentage. Run a few of your high-volume tickets through it and the answer stops being a guess.

The platforms won’t give you this math. Their reps quote the 15%. Your job is to know the 33%.

Quick reference: published commission rates #

• DoorDash: 15% Basic, 25% Plus, 30% Premier (DoorDash Merchant published rates 2024)
• Uber Eats: 15-30% Marketplace; flat 6% if you do your own delivery (Uber Eats published rates 2024)
• GrubHub: 15-25% commission + 10% delivery fee; all-in often 25-35% (GrubHub Marketplace rates 2024)
• Postmates: Merged into Uber Eats. Same structure.
• ChowNow: $99-149/month flat, 10% per delivery if using their drivers, 0% on dine-in pickups. Different model entirely.

Sources: DoorDash Merchant, Uber Eats, GrubHub Marketplace, Toast 2024 Delivery Economics Study, Restaurant Business Online, NRA 2024 Restaurant Industry Factbook.