Why Your Delivery App Orders Cost More Than You Think

Pull up your last delivery app payout statement and put it beside your P&L. Not the order-by-order screen — the statement. The gap between what customers paid and what landed in your account is the number that matters, and most owners have never converted that number into what it actually is: a line item bigger than the profit on the whole month.

This isn't an argument to delete the apps. They bring people who have never heard of you. That's worth something. But the way most Toronto restaurants use them — as the default place regulars reorder from — is the expensive version.

Commission isn't 30% of the order. It's most of the profit on the order.

Here's the trap in how we all talk about fees. "They take 30%" sounds like you keep 70%. You don't keep 70% of anything. You keep whatever is left after food, labour, packaging, rent, utilities, and card processing — and the commission comes out of that same thin slice, not out of the top.

Run the arithmetic on a $40 delivery ticket:

• Food cost at 30%: $12.00
• Packaging, bag, cutlery, label: $1.50
• Commission at 30%: $12.00

That's $25.50 gone before anyone has been paid to cook it. You have $14.50 left to cover labour, rent, hydro, gas, insurance, repairs, and the person answering the phone. In most independent kitchens those costs together eat 35–40% of sales. On a $40 ticket that's $14 to $16.

So the delivery order either breaks even or loses money. Not "earns less." Breaks even.

Now scale it to a month. Two hundred app orders at $40 is $8,000 in gross sales and roughly $2,400 handed over in commission. Ask yourself what your store actually netted last month. For a lot of independents in the GTA, that $2,400 is the same size as the net profit — or bigger. That's the reframe: commission isn't a marketing expense you pay out of profit. It is the profit, redirected.

Do this with your own numbers before you believe mine. Take last month's commission total off the statement. Write it next to your net profit for the same month. Those two figures side by side change how you feel about every decision that follows.

The costs that never appear on the statement

The commission percentage is the visible part. A few others stack on top:

Menu inflation you eat anyway. Most restaurants raise app menu prices to absorb the fee. Reasonable. But now your app price is 15–25% above your in-store price, and every regular who orders on the app is being taught that you're expensive. You pay for that later, in dine-in traffic.

Promo cost-sharing. "Free delivery over $30" and BOGO campaigns get pitched as platform marketing. Read the terms on who funds the discount. Often it's partly you, on top of commission.

In-app ad spend. You bid to appear higher in the same app that's already taking a cut of the order it sends you. It works, in the sense that it produces orders. It also means the cost of a delivery customer keeps climbing as more restaurants on your street bid for the same screen space.

Refunds and "order issues." A driver leaves a bag at the wrong door and the refund comes off your remittance, not theirs. You had no control over the last twenty minutes of that order and you're funding the apology.

No customer list. This is the one that compounds. You cook for someone forty times and still can't text them when you launch a new special. The relationship belongs to the app. You're renting your own regulars.

What the apps are actually good at

Discovery. That's it, and it's real.

Think of a delivery app the way you'd think of a stall in a mall food court. Foot traffic you didn't have to generate walks past. Somebody tries you for the first time because your photo looked good at 8pm on a Tuesday. You'd never turn that down.

The mistake is paying food-court rent on the customer who now specifically wants your banh mi and knows your name. That person doesn't need discovering. They need a faster, cheaper way to reach you — and if you don't build one, they'll keep using the expensive door because it's the only one they know.

So the goal isn't zero app orders. The goal is: first order on the app, second order direct.

How we did it at Banh Mi House

We run a restaurant too, and this is the part that took the longest to learn: nobody switches channels because of a sticker.

What moved the needle was the counter. When a regular came in for pickup, whoever was at the register said a version of the same thing every time: "You know you can order straight from us next time? Same prices as in here, and it comes right to our kitchen screen — no app in the middle." Then we handed them a card with the link and, for the first direct order, a small thank-you — a drink or a side, nothing dramatic. Staff were told to say it to anyone they recognized, not to everyone, because it lands differently when it sounds like a favour to a regular instead of a script.

The second piece was every bag going out the door. Not a coupon flyer — those get tossed. A small card, one sentence, our own ordering link, and the line "our prices, our kitchen, no third party." That reaches app customers too, which is the point: the app introduced them, the bag insert gave them a cheaper door for round two.

It was slow. Weeks, not days. The regulars went first, the once-a-month people took a few cycles. But the share of orders coming through our own system climbed steadily, and every one of those tickets carried its full margin instead of two-thirds of one.

Your own ordering has to be as easy as the app, or it won't work

This is where a lot of restaurants lose the plot. They put up an ordering page that takes six screens and doesn't remember the customer's card, then conclude that people prefer the apps.

People prefer whatever is fastest at 7pm with one hand free. Your system needs to:

• Work in a phone browser with no download and no account creation
• Remember the last order and the saved card
• Give an honest pickup or delivery time, not an optimistic one
• Text the customer when the food is ready
• Show the same prices as your printed menu, so switching feels like a reward
• Push orders straight to a kitchen screen or printer, so nobody is retyping tickets during a rush

If you want to see what that setup looks like end to end, that's most of what we build on our restaurant page — ordering, menu, and the phone side that catches the calls you miss.

A 60-day plan you can start on Monday

1. Week 1. Pull three months of app statements. Total the commission. Write it beside net profit for the same period. Now you know what you're solving.
2. Week 2. Get direct online ordering live with saved cards and honest timing. Match your in-store prices exactly.
3. Week 3. Print bag cards and train the counter line. One sentence, said to regulars, with a small first-order thank-you.
4. Weeks 4–8. Keep the apps running untouched. Don't cut them yet — you're testing whether the direct channel holds, not gambling the month's sales.
5. Day 60. Check the split: what share of orders came direct, and how many of those customers ordered more than once. Repeat rate is the number that tells you it's working.

Once a real share of your volume is direct, you have options you didn't have before. Raise app prices to properly cover the fee, since app customers are now first-timers rather than regulars. Turn off in-app ad spend and see if it hurts. Or just keep the commission you saved.

The apps stay useful for the job they're genuinely good at. You stop paying finder's fees on people you found years ago.