What a Delivery Marketplace Actually Costs
When UberEats or a similar marketplace brings you an order, the headline cost is the commission: a percentage of the order that goes to the platform. Delivery, promotions and "featured" placement are often layered on top as separate charges, and the exact terms vary by market, plan and what you negotiate.
It helps to think about that cost per order rather than as a monthly line item. A restaurant doesn't pay "the marketplace fee." It pays a slice of every ticket, including the large family order, the lunch-rush sandwich, and the regular who lives two blocks away.
Margins in food are thin, and a commission comes straight off the top of the sale, before food cost, labour and rent are covered. An order that looks healthy at the till can be barely profitable once the platform takes its share.
That doesn't make marketplaces a bad deal. It means you should know what each channel really costs you, and compare it to the alternative. Pull your last month of marketplace statements and work out three numbers:
- Average order value on the marketplace versus in your own channels.
- Total fees and commissions paid, including promotions you opted into.
- What's left after food and packaging costs for a typical order.
Most owners who do this exercise for the first time are surprised by the third number.
What You Give Up: The Customer Relationship
The commission is the visible cost. The less visible one is that the marketplace owns the customer, not you.
When someone orders from you in person, you can learn their name, ask for an email, notice they come in every Thursday. When they order through a marketplace, you generally get an order and a first name on a ticket. The platform keeps the contact details, the order history and the ability to reach that person again.
That has real consequences:
- You can't invite them back. A slow Tuesday is a problem you could solve with a message to your regulars. Without a list, you can't send one.
- You don't know who your best customers are. The person who orders every week looks identical to a one-time buyer.
- The platform can redirect them. The same app that brought a customer to you will show them your competitors the next time they open it.
- Your reputation is partly theirs. Ratings, search ranking and how your restaurant is presented are decided on their terms.
Think of a marketplace as renting a shop window on a busy street. The foot traffic is real, but the people walking past are the landlord's customers, not yours.
Tip: Before you change anything, count how many of your marketplace customers you could name or contact today. For most restaurants the answer is close to zero. That's the number you're trying to move.
When a Marketplace Still Makes Sense
It would be dishonest to pretend marketplaces don't earn their place. They do three things a small restaurant struggles to do alone:
- Discovery. People searching "pad thai near me" at 7pm on a Friday are already in buying mode. A new or little-known restaurant may never reach them otherwise.
- Delivery logistics. If you don't have drivers, a marketplace gives you a delivery fleet without hiring one.
- Low-effort volume. In a quiet season or a new location, extra orders from strangers can keep the kitchen busy.
So the realistic answer for most restaurants isn't "marketplace or own ordering." It's both, with different jobs:
| Channel | Best job | |---|---| | Marketplace | Finding new customers who've never heard of you | | Your own ordering link | Serving everyone who already knows you, and keeping them |
The goal is to use the marketplace as an introduction, then move people who like you to a channel where you keep the relationship and the margin.
What Switching Actually Takes
Moving toward your own ordering sounds like a big project, but for a single-location restaurant it's smaller than most owners expect. A realistic timeline looks like this:
Day one: build the menu. Write out items, sizes, modifiers and prices. If your menu is already on a marketplace, you've done most of this work. Be honest about prep times and any items that don't travel well.
Day one to two: set the rules. Decide opening hours for orders, pickup versus delivery, minimum order size, and how far ahead people can order. Keep it simple. You can add complexity later.
Day two to three: connect payments and test. Place a few orders yourself, walk them through the kitchen, and see where the ticket confuses people. Fix the labelling before real customers see it.
Week one: tell people. Put the link on your window, receipts, bags, social profiles and Google listing. Tell staff to mention it at the till.
Weeks two to eight: shift the habit. This is the slow part. Regulars won't move on their own. Give them a reason: a small perk, a stamp on their card, a menu price that reflects the saved commission, or simply a reminder that ordering direct supports the shop.
Two things to avoid. Don't switch off the marketplace on day one, because you'll lose orders while your own channel is still warming up. And check the marketplace's terms before promoting your direct link inside its packaging or app, since many have rules about it.
Why Owning the Checkout Compounds Over Time
A marketplace order is a one-time transaction. An order through your own checkout is the start of a relationship, and that's where the compounding happens.
Once you hold a customer's contact details and order history, a lot of things become possible that simply aren't on a marketplace:
- Repeat orders become predictable. You can see who orders weekly and who's drifted away.
- A loyalty program actually works. Stamps and rewards only make sense if you can recognise a returning customer. On a marketplace there's no way to tie visits together.
- Slow periods become fixable. A message to people who've ordered before costs nothing and often fills a quiet evening.
- Margins improve. The same order through your own channel keeps the share that would have gone to the platform.
None of this happens in month one. In month one, your direct channel will be small. But a restaurant that moves even a modest portion of its regulars across will, over a year, have a list of customers it owns, a set of repeat behaviours it understands, and a lower cost per order. A restaurant that stays fully marketplace-dependent will have the same dependence it started with, and a platform that sets the terms.
If you're a small independent, the most useful question isn't "should I leave UberEats?" It's "of the people who order from me, how many could I reach tomorrow without asking someone's permission?" Raise that number steadily and the rest follows.
Where to Start
If you want to try the direct route, set up your own online ordering page and give regulars one clear reason to use it. Shopolo's pricing is free to start with no monthly subscription, and you can sign up and build your menu in an afternoon without touching your marketplace listing.
Frequently asked questions
Is it worth paying UberEats commission on every order?
It depends on what the order is. For a first-time customer who would never have found you, the commission is a customer-acquisition cost and can be worth it. For a regular who would have ordered from you directly, it's money you didn't need to spend. The mistake is treating every marketplace order as equally valuable.
Can I stop using UberEats and just take orders myself?
You can, but most owners are better off running both for a while. Keep the marketplace for discovery, and steadily move regulars to your own ordering link. Drop the marketplace only if the numbers show it's costing more than it brings.
How do I get UberEats customers to order from me directly?
Put your own ordering link on your packaging, receipts, menu cards and social profiles, and make the direct option clearly better: a small perk, a stamp toward a reward, or a lower menu price. Check your marketplace's rules first, since many restrict how you can promote outside channels.
What do I lose by only selling through a delivery marketplace?
Mainly the customer relationship. You usually can't build your own email or phone list from those orders, can't invite people back, and can't see who your repeat customers are. The marketplace holds that data, and it can promote a competitor to the same people.
How long does it take to set up your own online ordering?
For a small restaurant with a menu that's already written down, expect an afternoon to build the menu and hours, plus a few days of test orders before telling customers. The slower part is changing habits: getting staff and regulars to use the new link.