Llama Studio All the work →
Sample case

Takeaway ordering that keeps the commission

Café and grill · sample built in five days

Hatch is not a real business. It is a working sample I built to show what this kind of site can do and the standard I build to. Every screen described below is live and you can use it. Nothing here is a client result, because there is no client.

Thirty per cent is the whole margin

A café doing four thousand pounds a month through a delivery marketplace hands over somewhere around twelve hundred of it. Not profit, revenue. For most independents that is more than the rent and roughly the entire margin on the food.

The trade you make is real: the marketplace brings you people who had never heard of you, and it handles the drivers. For genuinely new customers that can be worth paying for. The problem is the regulars. Someone who orders the same thing every Friday, who knows your name, who would happily order direct, still comes through the app and still costs you thirty per cent, every single week, forever.

The reason they use the app is not loyalty to it. It is that ordering directly from you is worse. A PDF menu, a phone number, and a conversation about whether you do the halloumi one without the chilli.

What I built instead

The whole menu, the options, the basket and the checkout, all on the café's own page. Twenty-six items with their real option sets, a slot chosen at checkout, collection or delivery, and the card payment handed to Stripe.

It opens on the food. There is no hero image to scroll past and no About Us before you can see what is for dinner.

01

The menu is the landing page

Someone arriving hungry at half past six wants to see food immediately. Every second spent scrolling past a photograph of the shopfront is a second in which the app is one tap away.

02

Options open in a sheet, not a new page

Choosing a size, a sauce and whether you want it without chilli happens over the menu rather than by navigating away. You never lose your place, which is the thing that makes people give up halfway down a long menu.

03

The basket is docked, not hidden

It sits at the bottom of the screen with the running total on it. No hunting for a cart icon in a corner, and no surprise at the end.

04

Collection or delivery is chosen at checkout, not before

Asking at the start forces a decision before anyone knows what it will cost or how long it will take. Asking at the end, alongside the time slot, matches how people actually decide.

05

Payment is handed over, and it says so

The card step goes to Stripe and is clearly labelled as such. Independents lose orders at exactly this point, because a payment page that looks improvised on a site nobody recognises is where people stop.

06

The kitchen gets an order, not an email to interpret

The confirmation is structured: items, options, time, method, reference. Whoever is on the pass should not have to decode a paragraph.

The part most people selling this leave out

Your own ordering page does not find you new customers. The marketplace genuinely does, and if a large share of your orders come from people discovering you for the first time then paying that commission may still be the right call. This is aimed squarely at the regulars, the ones who already know exactly what they want. Move those across and the commission you keep is the whole point. Do not let anyone sell you this as a replacement for discovery, because it is not one.

See it working

The sample is live and fully usable. Put a registration in, add things to a basket, go through the checkout. Nothing charges you and nothing is sent.