Short version: if you run more than 80 marketplace orders per month at an average ticket above $20, direct online ordering pays for itself inside 60 days. Below that you can still benefit, but the ROI window is 90-120 days. Above that, you're losing five figures a month to the 30% commission and you have no excuse.
The 30% nobody talks about honestly
Most operators know DoorDash takes "around 30%" but they think of it as a fuzzy operating cost. Let's get exact. The standard DoorDash partnership tier in 2025 is 30% commission on delivery orders and 15% on pickup. Uber Eats sits at 30%/15% as well. Grubhub is 20-30% on a tiered system. There are lower-commission tiers ("Marketplace Basic" at 15%), but they remove you from the promoted listings, so order volume drops by ~60%. Net effect: nobody actually uses the lower tier.
For a $40 delivery order on DoorDash, the math from the operator's side is roughly:
- Customer pays: ~$54 (food + delivery fee + service fee + tip)
- Restaurant gets paid: ~$28 (food revenue minus the 30%)
- Restaurant pays food cost: ~$11 (28% food cost)
- Restaurant's gross margin on that order: $17 on a $54 transaction. 31% gross margin.
The same $40 order placed direct on the restaurant's own site:
- Customer pays: ~$48 (food + delivery, no platform service fee)
- Restaurant gets paid: ~$38.80 (food revenue minus ~3% Stripe fee)
- Restaurant pays food cost: ~$11
- Restaurant pays delivery dispatch (DoorDash Drive, Uber Direct, or Roadie): ~$8
- Restaurant's gross margin on that order: $19.80. Plus the customer relationship is yours.
The break-even formula
Direct ordering software (like Zayos) costs roughly $399/mo + ~3% processing. The break-even point in marketplace orders saved is:
Break-even orders/mo = monthly_software_cost / (avg_ticket × (marketplace_commission − direct_processing_fee))
Plug in real numbers. Software $399/mo, avg ticket $35, marketplace commission 30%, direct processing 3%:
Break-even = $399 / ($35 × 0.27) = $399 / $9.45 = 42 orders/mo
Forty-two orders. That's one and a half a day. Past 42 orders/mo shifted from marketplace to direct, every additional direct order is pure margin recovery.
Three real operator scenarios
Scenario A. Small lunch spot, 600 orders/mo, $22 avg ticket
- Today: 80% marketplace at 30% = $3,168/mo in commission
- After: 25% of marketplace shifts to direct. Net new direct: 120 orders/mo. Saved commission on those: $792/mo
- Cost of Zayos: $399/mo
- Net month-one savings: $393. Net year-one savings: $4,716.
Worth it, but tight. At this volume the bigger payoff is the customer list — by month 12, your top 80 regulars are texting in their orders instead of clicking DoorDash.
Scenario B. Mid-volume neighborhood restaurant, 2,500 orders/mo, $34 avg ticket
- Today: 65% marketplace at 30% = $16,575/mo in commission
- After: 25% of marketplace shifts to direct = 406 orders/mo. Saved commission: $4,148/mo
- Cost of Zayos: $399/mo
- Net month-one savings: $3,749. Net year-one savings: $44,988.
This is the sweet spot. Most US independents fall here. Year-1 ROI on Zayos is roughly 11×.
Scenario C. High-volume / ghost kitchen, 8,000 orders/mo, $42 avg ticket
- Today: 70% marketplace at 30% = $70,560/mo in commission
- After: 25% of marketplace shifts to direct = 1,400 orders/mo. Saved commission: $15,876/mo
- Cost of Zayos: $399/mo (single location) or $799/mo (multi-brand)
- Net month-one savings: ~$15,000. Net year-one savings: ~$180,000.
Why most operators don't actually capture the savings
Three failure modes we see constantly:
- Direct ordering site exists but nobody knows about it. No QR code on the receipt, no SMS at checkout, no email follow-up. The site sits at 2% of orders for 18 months and the operator concludes "direct doesn't work for us."
- Kitchen can't handle multiple screens. The kitchen has a DoorDash tablet, an Uber Eats tablet, a Grubhub tablet, and now a fourth screen for the direct site. By week three the line cooks ignore the direct orders because the marketplace ones beep louder. Solution: one consolidated kitchen display. This is exactly why Zayos ships the direct ordering + Otter aggregator on a single tablet.
- No customer reactivation flow. A direct customer who hasn't ordered in 30 days needs a 15%-off SMS. Without that automation, your direct-order share plateaus at 8-10% and never grows.
The seven levers that actually shift orders
We covered the full list in the Uber Eats / DoorDash commission reduction deep-dive. The short version: QR codes on every printed surface, SMS list-building at checkout, one-tap reorder, direct-only specials, win-back automations, and a kitchen tablet that doesn't add screen number five.
What about commission-free marketplaces?
You'll see ChowNow, Slice, Toast Local, BentoBox, and others marketed as "commission-free." They charge a flat monthly fee instead — usually $99-$199/mo per channel. The math sometimes works for low-volume restaurants, but you give up two things: the discovery engine of being on DoorDash/Uber, and (in most cases) the customer data. Most of these platforms still keep the email and phone in their walled garden. Zayos and a few similar platforms give the operator full ownership of the customer record. Read the fine print on data ownership before signing anything.
How to decide
Two questions:
- How many marketplace orders do you do per month? Above 80, direct ordering pencils. Above 500, it's negligent not to.
- Will you actually drive customers to it? If you won't put QR codes on receipts, won't text the customer list, and won't run the win-back automations, don't bother. Direct ordering with no marketing is a $400/mo expense for zero benefit.
The 30-day shift target — what realistic ramp looks like
"How fast can I actually shift orders to direct?" is the question every operator asks. Realistic month-by-month ramp from a standing start:
- Month 1: 3-6% of total orders direct. Mostly your top 50 regulars who already know you.
- Month 2: 8-12%. QR codes have circulated, SMS list is past 200.
- Month 3: 14-18%. Win-back automation starts paying back. SMS list past 600.
- Month 6: 22-28%. The customer base has learned the lesson: "direct = cheaper / faster / specials".
- Month 12: 30-40% for operators who run the full motion. 18-22% for operators who launched the site but never marketed it.
The single biggest predictor of the year-12 number is whether the operator actually puts a QR code on every receipt. Sounds trivial. Isn't. Maybe 25% of operators we onboard actually do it consistently.
Hidden costs marketplace data won't show you
Three real costs of marketplace dependence that don't show up on the commission line:
- Refund absorption. When a marketplace refunds a customer for "missing item" or "cold food", you eat that cost — even when the customer is wrong. ~3-5% of marketplace revenue gets clawed back in refunds. Direct refunds, you control, so the rate drops to 0.4-0.8%.
- Menu tax. Marketplaces require you to charge ~12-18% more on their platform to recover the commission. Customers see those prices, compare them with your dine-in menu, and rate you "overpriced". The brand damage is real.
- Data lock-in. Two years of customer order history sit inside DoorDash's database, not yours. If their algorithm rotates you down or your account gets suspended, you have zero customer records to fall back on. Direct ordering builds the asset you actually own.
What the first 30 days of direct ordering actually look like in the kitchen
Operators imagine "going direct" as a marketing change. It's mostly an operational change. Here's the real first month on the line:
- Days 1-7: 2-4 direct orders a day. Mostly friends, regulars, and the test orders you placed yourself. Staff laughs at the new tablet beep.
- Days 8-14: 8-15 direct orders a day. QR codes are on receipts, SMS opt-ins start trickling in. Kitchen workflow tightens.
- Days 15-21: 18-30 direct orders a day. The win-back automation fires for the first time on lapsed customers and the email gets a 12-18% reorder click-through.
- Days 22-30: 30-50 direct orders a day. The single direct-vs-marketplace tablet on the line is now indispensable. Removing it would slow service.
Two things matter inside the kitchen during this ramp. First, the line cooks must trust that direct orders aren't going to overload them — start with capped order windows (max 4 direct orders per 10-minute slot) and relax as the team learns. Second, the manager has to enforce the QR-code-on-every-receipt rule. Without that one habit, direct order growth stalls at 3-4% and the whole exercise feels pointless.
Three operator profiles where direct ordering is non-negotiable
- The 90% marketplace operator. If marketplace is more than 80% of your volume, the day DoorDash changes its algorithm is the day your business dies. Direct is insurance, not just savings.
- The high-AOV concept. Average ticket over $45 (sushi, steakhouse, fine dining catering). The 30% commission is now $13.50/order — direct ordering hits payback inside 3 weeks.
- The brand-with-pull operator. If customers know your name, you've already paid for the brand. Letting DoorDash insert itself between you and that customer is leaving free money on the table.
FAQ
Will going direct hurt my marketplace visibility?
No. Marketplaces don't penalize restaurants for having their own site. We've tested this across 30+ restaurants — zero downranking, zero account warnings.
Who handles delivery for direct orders?
Zayos integrates with 14 dispatch providers (DoorDash Drive, Uber Direct, Roadie, Relay, etc.) and automatically picks the cheapest available driver. You don't hire a delivery fleet.
What if I'm pickup-only?
Even better. The math gets simpler — no delivery dispatch cost. Marketplace commission on pickup is 15%, processing is 3%. Break-even drops to ~30 orders/mo.
How long to launch?
2-3 weeks from contract to live. POS integration, menu mapping, QR code printing, and staff training are the only setup steps.
Run your numbers — Zayos · 15-min walkthrough on a real customer · phone 321-666-1102.