Delivery apps are eating your margin. Take your orders back.

Do the maths on last month's orders

Pull up last month's numbers from Uber Eats and Mr D. Add up every order that came through those apps, then apply a commission rate — commonly cited in the range of roughly 15–30%, depending on the platform and the plan you're on — to that total.

Say, hypothetically, a Garsfontein takeaway did R80,000 in delivery-app sales last month. Even at the lower end of that range, that's R12,000 gone before rent, stock, staff, or electricity get a look in. At the higher end, it's closer to R24,000. That's not a marketing cost. That's a slice of every single transaction, forever, for as long as the app is how the order arrives.

Run your own version of that sum before you read another word. It's the only number that matters in this post, and it's yours to calculate, not ours to guess at.

What you're actually paying for

To be fair to the platforms: you are getting something for that commission. Uber Eats and Mr D put your menu in front of people actively looking for food near Brooklyn, Hatfield, or wherever you are, and they run the driver network that gets it there hot. For a new customer who's never heard of you, that's a real service, and it's worth paying for.

The problem is who else is paying it. Your regulars — the people who already know your name, already have your food in their group chat, already decided you're their Friday-night order — are opening the same app out of habit and paying you through the same 15–30% cut. There was nothing to "find." They weren't discovered. They typed your name into a search bar because they already knew what they wanted. And the platform still takes its cut, same as if it had done the work of winning them over from scratch.

That's the part that adds up quietly. New-customer acquisition through an app is a cost of doing business. Paying full commission on a regular who orders from you every second Friday is a cost of not having anywhere else for them to go.

The part that should worry you more than the commission

Here's the bit that's easy to miss while you're staring at the commission line: the platform owns the relationship, not you.

When someone orders through Uber Eats, Uber Eats has their phone number, their order history, their delivery address, their favourite dish. You get an order number and a payout at the end of the week. You can't email them about a new menu item. You can't send a message when you launch a Moot Street pop-up. You don't even reliably know if the person ordering tonight is the same person who ordered three weeks ago — to you, they're both just an order ID.

Contrast that with a regular who calls you directly, or messages you on WhatsApp, or fills in an order form on your own site. That's your customer. Their number is in your phone, not locked inside someone else's app. If you're compared elsewhere — see our post on how customers search "restaurants near me" on Google — a direct relationship is also what lets you show up for that search on your own terms, not the platform's.

Commission is a cost you can calculate. Losing the relationship is a cost you only notice once a competitor with a cheaper delivery deal shows up on the same app, sitting right next to your listing.

Not "delete the apps" — a realistic split

None of this means cancelling Uber Eats and Mr D tomorrow. That would be a bad move for most restaurants, and we're not going to pretend otherwise.

The apps are genuinely good at one job: putting you in front of people who don't know you yet. Someone new to Waterkloof, someone traveling through, someone scrolling for something different on a Tuesday night — that's discovery, and it's worth the commission.

The regulars are a different job entirely. They don't need discovering. They need a reason and a way to order directly, so the second, third, and fiftieth order doesn't carry the same cut as the first one did.

The realistic split: keep the apps for new customers, build a direct channel for the people who already know you. Most restaurants running only on delivery apps have zero infrastructure for that second group. That's the gap.

What direct ordering actually needs

Direct ordering doesn't require a full ecommerce checkout system bolted onto your site — that's a bigger, more expensive build than most restaurants in Pretoria need. What it does need is three things, done properly:

A clear, easy-to-use online menu. Not a photographed page torn out of the printed one, not a PDF someone has to pinch-zoom on their phone. A real web menu, laid out for a screen, that loads fast on the kind of patchy 4G you get in a parking lot in Menlyn. We've written more on why a PDF menu quietly costs you orders compared to a proper web menu — worth a read if that's still how your menu lives online.

A phone or WhatsApp ordering flow that matches how South Africans actually order. This is not a foreign concept you have to teach customers — it's already how most regulars would rather order if the option is obvious and easy. A click-to-call button, a WhatsApp link that opens straight into a chat with your number, or a simple order-request form that lands in your inbox or phone — that's the infrastructure. It doesn't need to be complicated. It needs to work first time, every time.

A real reason for a regular to switch. Habit is powerful. If ordering through the app is what someone's always done, they'll keep doing it unless there's a reason not to. A direct-order-only special — a free side, a small discount, first access to a new dish — gives people the nudge to change a habit that's quietly costing you money every time they don't.

Practical ways to move regulars toward ordering direct

This is less about a big campaign and more about consistency in the small touchpoints regulars already see:

• A QR code on the till slip that goes straight to your WhatsApp order chat or web menu — the moment someone's paid once, make the next order easier.
• The same QR code and a short line ("order direct, skip the app fee, get 10% off") on takeaway packaging and delivery boxes — it travels home with every order, app or not.
• A pinned post and regular mentions on your social pages pointing at direct ordering, not just food photos.
• A flyer insert in delivery bags for the first few weeks of a push — cheap, physical, and it's in someone's hand at exactly the moment they're thinking about your food.

None of this needs to replace the apps or compete loudly with them. It just needs to exist, consistently, so regulars have somewhere else to go.

The honest cost comparison

Here's the framing that actually matters when you weigh this up. A website built to handle this properly — a clear menu, a working phone/WhatsApp ordering flow, a page built for how people search for you locally — is a one-off cost. At JLC Studio, a Launch package is R10,000 to build with a R700 monthly retainer for upkeep. Growth, with more pages and more room to grow the site over time, is R16,500 build with a R950 retainer. Neither of those numbers moves no matter how many orders come through the door.

Commission moves with every single order, indefinitely. R700 a month is roughly what a Garsfontein-sized restaurant might lose in commission on one decent weekend of app orders from existing regulars. The website is a fixed cost. The app fee is a percentage tax on your own success — the busier you get, the more it takes.

That's not an argument against delivery apps. It's an argument for not paying full app commission on customers who were never the app's to find in the first place. If you want to see how we approach this for restaurants and food businesses specifically, that's the starting point — a site built to carry its weight on both fronts, discovery and direct ordering, not just a menu online for the sake of it.

Let's price it against your commission bill.