Does Your Restaurant Need Its Own App? The Numbers
By CodexierPublished 6 min read
A restaurant app sounds like the modern thing to have, and app vendors are happy to sell one. The honest question is whether your customers would install it, and whether the same result can be had from a web page that needs no installation. This guide runs the numbers and gives a rule for the few cases where an app actually wins.
What a restaurant app would do
Concretely, a restaurant app covers ordering for pickup or delivery, saved payment methods, order history and re-order, a loyalty programme, push notifications for offers, and sometimes table booking. Every one of those except push notifications works equally well on a web page saved to the home screen. Push is the genuine app advantage, and it only matters if customers want to hear from you between visits.
Build and yearly costs
| Cost line | App | Web ordering page |
|---|---|---|
| Build | Fixed build price, two platforms from one codebase | A fraction of the app build; part of an e-commerce setup |
| Store accounts | Yearly Apple fee and a one-time Google fee, in your company's name | None |
| Updates | Both stores change requirements yearly; an app that is not updated stops working within a couple of years | The platform updates itself |
| Backend and hosting | Needed for both | Needed for both |
| Payment fees | Same providers (Swish, card); physical goods are exempt from store commission | Same |
| Marketing the channel | You must drive installs | You must drive visits, but a link is one tap, not a download |
The build price is the visible number. The yearly update line is the one that decides whether the app is still alive in year three.
The mechanism behind the update cost is worth understanding: Apple and Google change their operating systems every year, and apps that are not rebuilt against the new versions eventually get warnings, then removal. A web page has no such clock. Our app maintenance plan exists precisely because this line is unavoidable, and it should be in your calculation from day one.
Downloads: the hard part
A customer who wants to order from you on the web taps a link and is ordering in seconds. A customer who wants to use your app has to find it in the store, download it, open it, create an account, and only then order. Every step loses people, and the biggest loss is the first: most customers of a single restaurant will not install an app for it, because they already have the delivery apps and the phone has limited attention. An app is a channel you have to fill, and filling it costs promotion at every table and on every receipt for months.
- Installs come from your own promotion, not from the app store; nobody browses the store for a local pizzeria.
- Every install must beat the delivery apps already on the phone on convenience, price or loyalty.
- Uninstall rates for single-business apps are high; the app must be worth keeping between visits.
- A web page saved to the home screen looks and feels like an app to most customers, without the store.
Web ordering as the alternative
A mobile-first ordering page on your own domain gives you the margin advantage over delivery apps, the customer data, saved orders and a loyalty stamp, with none of the install friction. Built as a progressive web app it can be added to the home screen with your icon, work offline for the menu, and even send limited push notifications on Android and recent iPhones. We covered the setup in detail in online ordering without app fees.
Start on the web
Ordering, Swish and card, time slots, loyalty. Measure how many customers order direct each month.
Add home-screen install
A small banner inviting regulars to add the page to their home screen. Free, and it tells you whether people want an app-like relationship.
Then decide on an app
If hundreds of customers order monthly and push notifications would change their behaviour, the app case is real. Build it on the same backend.
When an app does pay off
Three cases hold up. A chain with several locations where one app serves all of them and the marketing cost is shared. A high-frequency concept (coffee, lunch, bubble tea) where the same customer buys several times a week and a wallet plus push offers measurably increases frequency. And a subscription or membership model where the app is the membership card. In each, the customer has a reason to keep the app between visits, which is the whole question.
If you are in one of those cases, the build is a fixed-price project with us and the numbers are on the pricing page. If you are not, we will say so on the intro call and point you to the web version instead; a fifteen-minute honest answer is cheaper for both of us than an app nobody installs.
- Cross-Platform MVP App BuildOne codebase for iPhone and Android, published in both stores, on a backend you can share with your website.
- App Maintenance & ScalingThe yearly line: OS updates, store compliance, monitoring and small changes.
- Book a free 15-minute callTell us your monthly order count and how many are regulars; we say plainly whether an app or a web page is the right build.
Frequently asked questions
Do Apple and Google take a commission on food orders in an app?
No. Store commission applies to digital goods and subscriptions inside the app, not to physical goods and services delivered outside it, so a restaurant app can use Swish, card or Klarna without paying a store share. You do pay the developer account fees and any payment provider fees as usual.
Can a web page send push notifications like an app?
On Android, yes, once the user adds the page to the home screen and allows notifications. On recent iPhones it also works after the page is added to the home screen. The reach is smaller than a native app's and the customer has to opt in, but for a single restaurant it is often enough to test whether notifications change anything.
What does an app cost per year to keep alive?
Count the developer accounts, hosting for the backend, and at least one update cycle per year to stay compatible with new iOS and Android versions, plus any changes you want. The exact figure depends on the app, which is why we run a maintenance plan as a fixed monthly package rather than quoting a generic number here.
We have several locations. Does that change the answer?
It strengthens the app case, because one app serves every location and the marketing cost is shared, and because customers of a chain are more likely to keep the app. Even so, start by measuring direct orders on the web; a chain with low repeat frequency still gains little from an app.
App or web ordering? Bring your order numbers
In fifteen minutes we run the arithmetic on your monthly orders, regulars and locations, and tell you honestly which build makes sense and what it would cost, including the yearly costs.
Book a free 15-minute call