The two commissions eating your restaurant: how to work out what you really pay

There are two different commissions eating your margin. One charges you for every order. The other, for every person who sits down. Almost nobody separates them, which is why almost nobody knows what they really pay.

Ask a restaurant owner what "the platforms" cost them and you'll get a round number, from memory, with a look of annoyance. Ask which of the two commissions hurts more and they go quiet.

They aren't the same thing. They work differently, they're paid differently and you get rid of them differently. This article separates the two and gives you the formula to work out yours with your own numbers — not with a table someone invented on the internet.

The commission per order

This is the delivery aggregators'. You cook, they bring the order and they take a percentage of every ticket, calculated on the total, not on your profit.

That's the trap almost nobody gets right: the percentage doesn't eat a slice of your revenue, it eats an enormous slice of your margin. If your real margin on a €30 order is €9, a 30% commission on the total is €9. In other words: the whole order for them and the whole job for you.

Multiply that by the orders of an August weekend and you'll see why some kitchens work at full capacity and close the month with nothing left.

The commission per diner

This is the booking platforms', and it's the one almost nobody has counted. They don't charge per ticket: they charge for every person who sits at your table having come through their site.

The expensive part here isn't the fee. It's that the customer isn't yours. They book on the platform, get the reminder from the platform, leave the review on the platform, and next time they go in through the same door — paying again. Their name, their phone number and their visit history live in a database you have no access to.

A customer who comes back four times a year costs you four times. Forever.

How to calculate yours (and why the tables online are useless)

You'll find articles with very specific pricing tables. Distrust all of them: rates change with the plan, the city, the volume and whatever you negotiated the day you signed. Your contract is the only valid data. Get it out, and fill this in:

Item	Where to find it	Example
Average ticket	Your POS	€30
Orders per month via aggregator	Their billing panel	120
% commission per order	Your contract	30%
Diners per month via booking platform	Their panel	90
Fee per diner	Your contract or your invoice	Yours
Fixed fee or visibility plan	Your monthly invoice	Yours

Then apply two multiplications that aren't remotely difficult:

• Annual cost of delivery = average ticket × orders per month × % commission × 12
• Annual cost of bookings = diners per month × fee per diner × 12, plus any fixed fees

With the example above, that first line alone is €12,960 a year. And it's half the bill.

If you can't be bothered with the calculator, our restaurants page has both sums built in with sliders: move your numbers and see the annual total.

What you are actually buying when you pay

This is where most articles on the subject lie by omission, so let's say it: the platforms do something you don't.

They bring demand you don't have. A tourist who arrives on a Tuesday in a town they don't know opens the app already on their phone; they don't search for your website. A listing with two hundred reviews gives them a confidence your new site can't give them yet. And the payment, the confirmation and the reminder all work without anyone on your team touching a thing.

That's worth money. The question isn't whether it's worth money: it's whether it's worth all the money you're paying, for every customer, including the ones who already know you and would have come anyway.

Because that's where the real waste is. Paying for a new customer is acquisition. Paying for the neighbour who's come in every fortnight for three years is a tax.

Why a website with a WhatsApp button doesn't replace them

This is the mistake people make trying to fix it cheaply: build a nice landing page with a "book on WhatsApp" button and expect people to stop using the app.

They don't. And it isn't the customer's fault: it's that the button doesn't compete. For an owned channel to genuinely replace a platform it has to do, at minimum, what the platform does:

• Confirm on the spot. If the customer has to wait for someone to read the message, they haven't booked: they've asked.
• Know how many seats are left. With no real capacity per sitting you eat the overbooking, and one table too many on a Saturday costs more than a whole month of commission.
• Alert you without you watching anything. The booking has to ring on the restaurant's phone, not sit waiting in a dashboard.
• Be cancellable by the customer. A cancellation link avoids the empty table nobody warned you about.
• Keep the customer. Name, phone and history in your database. This is what turns the system into an asset instead of a cost.
• Let people find you. If your menu lives in a PDF, Google can't read it and you don't show up when someone searches for what to eat nearby. We cover that in how to get AI to recommend your restaurant.

When the owned channel does these six things, the platform stops being essential and becomes one more source of new customers — which is exactly the role it should have.

When you should NOT drop them

With the same honesty: if your own demand today is zero, cancelling the platforms doesn't save you money. It takes diners away.

The order that works is this one, and it's boring:

1. Build your own channel and get it genuinely working.
2. Start sending the people who already know you there: the table QR, your social profile, the card with the bill, your Google listing.
3. Measure what percentage of bookings comes in through your own house each month.
4. When that percentage is solid, renegotiate or downgrade the platform plan — not before.

Nobody gets rid of a commission by cancelling. You get rid of it by building the alternative first.

Two restaurants on the same street

In Puerto Pollença we have two clients on the same street, and between them you can see the whole problem.

Tirano Street Food is the case of someone paying neither: a sit-down restaurant, no delivery and no portal. Digital menu, direct bookings over WhatsApp and a site tourists find by searching — in two and a half months, 303 clicks from Google and three in ten arriving from outside Spain.

Tierra de Fuegos is the per-diner one: their own slot-based booking system, with automatic confirmation, an instant alert to the restaurant and every diner's details in their own database. The brief was never "have a website": it was to stop paying for every person who sits down.

Same street, same town, two different problems, the same underlying remedy: the owned channel has to be as good as theirs.

---

Want to know what they cost you? Get your contract out, do the two multiplications above and compare the result with what having your own system costs. If the number surprises you, get in touch and we'll look at it together.