Delivery-app commissions vs your own website: the break-even math nobody shows you

The number that actually leaves your pocket

Most owners know the delivery apps take a cut, but the sticker commission understates it. A marketplace order on DoorDash or Uber Eats is rarely a single 15% line. There is a base commission tier, then a separate delivery cost, a service fee, and often an optional marketing or 'sponsored listing' spend that many restaurants feel pressured to buy just to stay visible. Stacked together, the effective take on a delivered order commonly lands in the 25-30% range.

Run it on a real ticket. A customer orders $40 of food. At a 28% effective take, roughly $11.20 goes to the app before you have paid for a single ingredient. Your food cost on that order might be another $12. What looked like a $40 sale is now feeding you maybe $16 of gross margin — and you did the cooking, the packing, and the customer service. The app did the click.

What a direct order costs instead

When a customer orders through a page on your own website, the money flows differently. You are not renting a marketplace; you are taking a card payment. The only unavoidable cost is the payment processor — Stripe, Square, or whatever your online-ordering tool uses — which runs in the neighborhood of 2.9% plus a fixed cents-per-transaction fee. On that same $40 order, that is a little over a dollar, versus eleven.

That gap is the whole argument. You are trading a percentage that scales with your success for a flat fee that barely moves. Every dollar of difference is margin you keep — the difference between an order that barely clears food cost and one that actually contributes to rent, labor, and profit. Nothing about the food, the packing, or the driving changes; only who gets paid for the transaction.

Doing the break-even for your restaurant

The honest question is not 'is direct cheaper' — it obviously is per order — but 'how many orders do I need to move to make the channel worth building.' That is a break-even calculation, and it is simple enough to do on a napkin.

Say a direct order saves you roughly $8 in avoided commission versus the app on your average ticket. If building and hosting your ordering channel costs on the order of a couple hundred dollars a year, you break even once you have shifted about 25-30 orders across the whole year to direct — less than one a week. Everything past that point is recovered margin that used to belong to the app.

• Find your effective app take from a recent payout statement, not the advertised rate.
• Multiply it by your average order value to get dollars-lost-per-order.
• Subtract your direct-order processing cost (about 3%) to get dollars-saved-per-order.
• Divide your annual channel cost by that savings — the result is how many orders you must move to break even.

This is not 'quit the apps' — it's stop overpaying

The apps genuinely do one thing well: discovery. A hungry person browsing DoorDash with no particular restaurant in mind is a new customer you might never have reached, and paying a commission to acquire them can be worth it. The waste is paying that same acquisition fee over and over on people who already know you, already love your food, and would happily order direct if you gave them an easy way.

So the strategy is a split, not a boycott. Keep the apps as a paid discovery channel for new diners. Then give your regulars — the ones who order the same Friday-night dinner every week — a direct path that costs you 3% instead of 28%. You are not fighting the marketplace; you are refusing to rent your own repeat customers back from it every single week.

The compounding cost of app-only regulars

Here is the part that quietly hurts the most. A loyal customer who orders from you every week through the app is not a one-time 28% hit — it is 28% fifty-two times a year, forever. That single relationship might represent well over a thousand dollars of food sales annually, and a big slice of it evaporates in commission on orders that were never at risk of going anywhere else.

Move even a fraction of your repeat volume to direct and the recovered margin does not just add up, it compounds, because those are your highest-frequency customers. The apps are expensive precisely where they are least necessary: on the people who already chose you. A direct channel is how you stop paying a finder's fee for customers you found long ago.

What the direct channel needs to actually work

A break-even only matters if customers actually use the channel, so it has to be genuinely easy — not a clunky afterthought that sends people back to the app out of frustration. The direct-ordering experience has to feel as smooth as the marketplace they are used to, or the math never gets a chance to play out.

In practice that means a mobile-first ordering page, a menu that is fast to browse, saved payment, clear pickup and delivery options, and an order confirmation that inspires confidence. Get those right and customers happily switch, because ordering direct is often a little cheaper for them too — no marked-up menu prices and no service fee. The next guide in this hub walks through setting that channel up end to end.