Payments for Ghost Kitchens: What DoorDash, Uber Eats, Stripe, and Square Really Cost You

The short version: a ghost kitchen’s “payment processing” is really two businesses — the delivery apps taking 15–30% and your own ordering at ~3%. On $60,000 a month in volume, the apps take $9,000 to $18,000; your own site would take about $2,000. The apps are not your processor; they are your landlord. This guide does the math the apps and the big processors will not: every commission tier computed, the own-ordering comparison, payout reconciliation, and the honest verdict on each rail. None of them will ever write “payments for ghost kitchens.” We just did.

Rates and commission tiers checked October 2026.

The one number: the apps as a second rent

4 brands, $60,000 a month in delivery-app volume, blended 25% commission = $15,000 a month, $180,000 a year paid to DoorDash, Uber Eats, and Grubhub. Your kitchen rent is probably $5,000–$8,000 a month. The apps cost roughly twice your rent — and unlike the landlord, they also own the customer relationship, the data, and the reorder. The entire payments strategy for a ghost kitchen is one sentence: use the apps for acquisition, convert to your own ordering for margin. Everything below prices the pieces of that sentence.

The commission table: every app tier, computed

Published 2026 plan tiers, computed on $60,000/month in volume (per-transaction math uses a $22 average ticket):

• DoorDash Basic (15% delivery): $9,000/month — limited visibility and delivery radius. For an unknown brand, cheap and invisible.
• DoorDash Plus (25%): $15,000/month — DashPass eligibility, the tier most operators actually live on.
• DoorDash Premier (30%): $18,000/month — best placement plus a growth guarantee that refunds the month under 20 orders. Read the guarantee’s fine print; it is narrower than the sales call.
• Uber Eats Lite (~15–20%): $9,000–$12,000/month — basic visibility, search-only discovery.
• Uber Eats Plus (25%, 30% on Uber One orders): $15,000–$18,000/month.
• Uber Eats Premium (30%): $18,000/month — Uber One priority placement.
• Grubhub Basic (5% marketing + 10% delivery + $0.30 + 3.05% processing): roughly $10,800–$12,000/month all-in on delivered orders.
• Pickup orders: DoorDash 6%, Uber Eats 6–10% — the cheapest app orders you will ever get, and the reason your direct site should push pickup hard.

Two things jump out. First, the tier spread is $9,000 a month — Basic to Premier is the difference between a profitable kitchen and a busy one. Second, none of these tiers include the extras: sponsored placement, promo participation pressure, error adjustments, and the menu-price markup you need (10–20% over direct) to survive them. The headline commission is the floor, not the ceiling.

Your own ordering: the 3% rail, priced

Now the other side — what direct ordering actually costs on $60,000/month:

• Stripe (2.9% + $0.30 on ~$22 tickets): ~$2,110/month. Payment links, Apple Pay, card-on-file — the full toolkit.
• Square online (3.3% + $0.30): ~$2,370/month — the quiet tax on every direct order if your POS is Square.
• DoorDash Storefront (2.9% + $0.30): ~$2,110/month — DoorDash’s own commission-free ordering widget. Yes, the irony is noted: the cheapest way to escape DoorDash’s commission is DoorDash’s other product.
• Delivery for direct orders (DoorDash Drive / Uber Direct): flat per-delivery fees (~$6–$10) instead of commissions — the piece that makes direct delivery work without your own drivers.

The math: moving $20,000 a month from Uber Eats Plus (25% = $5,000) to your own site (~$700 in processing + ~$400 in delivery fees) saves roughly $3,900 a month. That single shift funds the entire tech stack in our tech guide several times over. The apps will not tell you this; their business model is the gap.

The payout reconciliation nobody does (and the money in it)

Three apps × four brands = twelve payout streams with commissions, refunds, adjustments, promo deductions, and error charges — each on its own statement format, each on its own schedule. The industry data point from our research files: a ghost kitchen that consolidated four brands’ payouts into one weekly reconciliation found $800 a month in app-fee errors. That is not unusual; it is normal. The apps’ statements are built to be paid, not audited.

• Weekly, not monthly. Reconcile expected vs. actual payouts every week — commission applied vs. commission tier, refunds you actually approved vs. refunds charged, promo deductions with your sign-off vs. without.
• Dispute in writing, fast. Every app has a dispute window; miss it and the adjustment stands. The reconciliation routine should produce the dispute list the same day.
• The card-on-file pattern for direct. Your own site’s repeat customers should check out with stored cards (the pattern in Finix’s public docs) — one-tap reorder is what makes direct competitive with the app’s one-tap habit.

Flat vs. interchange-plus — and why it barely matters here

On your own ordering, the flat-vs-interchange-plus debate is a rounding error next to the commission question: Stripe’s 2.9% + $0.30 vs. interchange-plus at ~2.2% + $0.25 saves roughly $400 a month on $60,000 — against $13,000 in commission savings from moving volume direct. Do not optimize the 3% rail; migrate volume onto it. The one exception: if your direct site ever passes $100,000 a month, get the interchange-plus quote — at that point the $700/month savings is worth the paperwork.

The zero-fee angle the big processors never mention applies sideways here: the apps are the fee. There is no cash-discount play against a 25% commission — the play is the direct rail itself. The big processors cannot sell you this either; your direct ordering at 2.9% is their smallest-margin business. Everyone’s incentives point at keeping you on the apps. Yours should not.

The honest verdict on each rail

• DoorDash: the acquisition channel. Largest driver network, best POS integrations, three honest tiers. Use Basic or Plus for discovery; treat Premier’s 30% as a marketing budget with a P&L, not a default. Never run all four brands on Premier — that is $18,000 a month in rent for customers you do not own.
• Uber Eats: the urban channel. Strong in dense cities, Uber One orders at 30% on Plus. Same rule: acquisition, not infrastructure.
• Grubhub: the corporate channel. Smaller, but Grubhub Corporate orders are the catering-adjacent business — price it as lead gen for your bulk-order operation (see the AI guide’s voice workflow).
• Stripe: the direct rail. 2.9% + $0.30, payment links, invoicing, ACH at 0.8% capped at $5 for the corporate bulk orders. The boring backbone of the 3% rail.
• Square: fine for the counter you do not have. If you run a pickup window, Square’s free plan works; online, its 3.3% + $0.30 is the most expensive direct rail on this page. Do not default to it.
• Chase, Global Payments, Worldpay, Fiserv, Adyen: pass. Nothing here that moves the commission needle — which is the only needle.

The Finix angle: operators running real direct volume need card-on-file, invoicing, scheduled charges, and ACH in one stack — the plumbing that makes the 3% rail actually convert. The apps will never build you an off-ramp; build your own.

Do this in order

• This week: audit which tier each brand is actually on — operators drift onto Premier and never notice. Drop the brands that do not need the visibility back to Plus or Basic.
• This month: launch (or fix) your own direct ordering with first-order discounts, and put the URL in every bag. Start the weekly payout reconciliation.
• This quarter: measure direct-order share monthly with a target: 20% of volume direct within two quarters. Every point of share shifted is ~$220 a month back in your pocket.

Payment processing is the only vendor that charges you more every time you succeed — and in ghost kitchens, the vendor is the app. Treat commissions like rent: negotiate the tier, audit the statements, and build the direct rail that makes the landlord optional. On the direct rail, corporate bulk orders still need invoices — ClaroPro’s voice-to-invoice speaks the order into an invoice with a payment link, no back office required.

Next: AI for ghost kitchens, the ghost kitchen tech stack, and the ghost kitchens hub.