Support

Lesson 1 of 11 ·10 min read ·Reviewed 5 August 2026

Web Apps vs Mobile Apps

What the differences are, why fast food chains build what they build, and how to decide for your own product.

1. The core distinction

A web app runs in the browser. You host it on a server, users reach it through a URL, and you control everything about how it deploys and updates. This site, endpoint51.com, is exactly this model.

A mobile app (the kind on Google Play or Apple's App Store) is installed on the user's device. It is packaged, signed, reviewed by the store, and updated on the user's schedule, not yours. The industry word for this kind of installed app is a native app; you will meet that term everywhere in mobile development, and this lesson uses it from here on.

One distinction matters from the start: installed does not automatically mean platform-native. An app you install from a store might be written directly for that platform (Swift, Kotlin), built with a cross-platform framework (Flutter, React Native), or built largely from web code inside a native container. The user installs all three the same way and cannot tell them apart from the icon.

And from the server's perspective, every one of them is an API client. A web app, an iPhone app, and an Android app can authenticate against the same backend and request the same resources. What changes between them is distribution, device access, storage, and release cycle, not the API behind them. That idea runs through this whole course.

2. Key differences at a glance

Web appGoogle Play / App Store app
Deployment Push to server; everyone gets the new version instantly Upload to store, pass review (hours to days), users update whenever their device does
Gatekeeper None. Nobody can reject or remove your site Google/Apple review every release and can remove your app. Play requires a $25 one-time fee; Apple charges $99/year
Payments (digital goods) Stripe or similar, roughly 3% per transaction Store billing by default, with a 15–30% cut; court and regulator pressure is opening external-payment exceptions (covered in Payments in Apps)
Payments (physical goods) Your own processor, ~3% Also your own processor, ~3%. The store cut does not apply to physical goods like food
Push notifications Limited and patchy on mobile, historically weak on iOS Reliable, first-class
Device access Narrow, permission-gated (some location, camera, storage) Deep: sensors, Bluetooth, background work, file system, offline by default
Presence A tab or bookmark Home-screen icon; much stronger habit and re-engagement loop
Codebase One codebase (HTML/CSS/JS + backend) for every platform Kotlin/Java for Android, Swift for iOS, or a cross-platform framework (Flutter, React Native) covering both
Maintenance One deployment pipeline; the browser absorbs device differences Two build toolchains (Xcode, Android Studio), yearly OS releases deprecating APIs, signing certificates, and store SDK requirements to track
Discovery Search engines, links, SEO Store search and rankings, a separate and very competitive game
Versioning Everyone is always on the latest version You always have users on old versions you must keep supporting

Play Store note for solo developers: new personal Google Play accounts must run a closed test with 12+ testers before publishing publicly. This trips up a lot of independent developers, so budget time for it if you ever go that route.

3. The middle ground: PWAs and wrappers

Web apps and native apps are not the only options. Several approaches let you reuse web technology while adding installation, app-store distribution, or deeper access to device features.

Progressive Web App (PWA)

A web app with a manifest (name, icon, installed appearance) and a service worker (caching, offline behaviour) so it can be added to the home screen, work partly offline, and send push notifications. Still fundamentally a web app reached by URL, and capabilities vary by browser and OS: on iOS, web push only works after the user has added the app to their home screen.

Trusted Web Activity (TWA)

A thin Android shell that shows your live website full-screen through the user's browser engine, published on Google Play. Domain verification proves the site and app belong to the same developer (the "trusted" part), and nothing is bundled inside, which is what separates it from a WebView wrapper. The cheapest way to get a web product onto Play; Google's billing and quality rules still apply to what is inside.

Hybrid wrappers (Capacitor, Cordova)

Your web code ships inside a native container, rendered by an embedded browser component, with plugins bridging JavaScript to native features (camera, notifications, files, haptics). More device access than a PWA or TWA, still one codebase, and a good fit for web-skilled teams. The catch: the result is a real store app you must build, sign, submit, and maintain.

App Store note for wrappers: the two stores treat this middle ground very differently. Google Play accepts TWAs and thin wrappers. Apple does not ban hybrid apps outright, but App Review Guideline 4.2 (Minimum Functionality) asks for features that "elevate it beyond a repackaged website", and reviewers regularly reject Capacitor or Cordova shells that would sail onto Play. If the App Store matters to your product, plan real native functionality, not just a wrapper.

Cross-platform native (Flutter, React Native)

One codebase compiled to genuinely native apps for both Android and iOS. Not web tech at runtime (Flutter uses Dart, React Native renders native widgets). This is what most mid-size and large consumer apps use today.

4. What fast food takeaways use, and why

Big chains (McDonald's, Burger King, Domino's, KFC) almost all run native store-distributed apps, usually built with a cross-platform framework, plus a web ordering site alongside. The reasoning maps directly onto the table above:

  • Push notifications are the business model. "20% off today only" and loyalty nudges need reliable native push. This alone justifies the native app.
  • Loyalty and habit. The home-screen icon, saved usual order, and points balance drive reorder frequency in a way a browser bookmark never will.
  • The store tax does not hit them. Food is a physical good, so the 15–30% store cut on digital purchases does not apply. They process payments with their own provider (Stripe, Adyen) inside a store-distributed app and pay normal card rates. They get store distribution without the store tax.
  • Location features. Store finders, geofenced offers, and order-ready timing all benefit from native location access.
  • Their marketing links open the app. Offer emails and discount links use Android App Links / iOS Universal Links: ordinary https links that open the installed app directly at the offer, and fall back to the website when the app is not there. One link serves both halves of the funnel.
  • The web version catches everyone else. First-time and low-commitment customers order in the browser with nothing to install; regulars get funnelled into the app where the loyalty flywheel lives.

Small independent takeaways

They typically build nothing themselves. Two common routes:

  • Marketplace apps: list on Just Eat, Uber Eats, or Deliveroo. Zero build cost, but commissions run roughly 14–30% per order and the platform owns the customer relationship.
  • White-label platforms: services like Flipdish (Irish), Storekit, or Slerp generate a branded app and ordering website for a monthly fee plus a small per-order cut. The takeaway keeps the customer data and pays far less per order than marketplace commission.

Many independents use both: marketplaces for discovery, their own white-label channel for repeat customers, often with a "order direct and save" incentive to shift regulars off the commission-heavy marketplaces.

5. What web-first software companies use, and why

Flip the example: Figma, Canva, Linear, and Basecamp all built their products web-first, and for desktop use most of them are web-only or close to it. Their reasoning reads like the same table with the weights reversed:

  • Every user is on today's version. A fix or feature deploys to the server and reaches everyone on the next page load. No review queue, no waiting for users to update, no old client versions to keep supporting.
  • A URL is the onboarding. Sharing a design or a project means sending a link; the recipient clicks and is inside the full product in seconds, with nothing to install. For collaboration software, that loop is the growth engine.
  • The store tax would hit them. These are digital subscriptions, exactly the category the 15–30% cut applies to. Selling seats through their own web checkout at ~3% is the same margin calculus as the decision table below, made at company scale.
  • The browser ceiling is higher than it looks. Figma runs a C++ rendering engine compiled to WebAssembly on a GPU-accelerated canvas, in a tab. Most dashboard-and-documents products never get near the sandbox's limits.

The desktop app that is secretly the web app

When these companies do ship a "desktop app" (Slack, VS Code, Figma), it is usually the same web code inside an Electron shell (or a lighter equivalent like Tauri): a native window, system tray, and global keyboard shortcuts around a product that stays one codebase. It is the wrapper move from section 3, pointed at the desktop instead of the app stores.

6. How to decide for your own product

If your product is…Build…Because…
Content, courses, or paid digital access (like an online book) Web only Instant deploys, Stripe's ~3% instead of the store's 15–30%, full SEO, and no gatekeeper who can remove you
Forms-and-data / dashboard style SaaS Web first, PWA if mobile use grows One codebase, always current, and desktop is usually the main surface anyway
Something driven by notifications, habit, or loyalty Native (cross-platform), web alongside Push and home-screen presence are the levers, and they need native
Something needing hardware: sensors, Bluetooth, heavy offline Native The browser sandbox simply does not reach far enough
Selling physical goods with a mobile-heavy audience Web first; add native when repeat-purchase volume justifies it The store cut does not apply to physical goods, so the native app is a retention tool, not a requirement

Rule of thumb: the web is the default. Go native when one of the native-only capabilities (push, hardware, offline, home-screen habit) is central to how the product wins, not merely nice to have. And never because an app simply feels more substantial. This lesson explains what separates the two; when you are weighing the call for a specific product, the decision guide Which Should You Build First? walks you through making the choice. And if you sell digital content, remember the store tax: moving a $69 sale from Stripe to store billing turns a roughly $2 fee into a roughly $10–21 one.

7. Quick glossary

TermMeaning
Native appAn app installed from a store. It may be platform-native (Swift on iOS, Kotlin on Android) or compiled for the platform by a cross-platform framework
Platform-native appAn installed app built directly with the platform's own tools: Swift/SwiftUI on Apple platforms, Kotlin/Jetpack Compose on Android
PWAProgressive Web App: an installable, offline-capable web app, no store needed
TWATrusted Web Activity: an Android shell that publishes your live website on Google Play
Hybrid appWeb code shipped inside a native container (Capacitor, Cordova)
Cross-platform frameworkOne codebase producing native apps for both Android and iOS (Flutter, React Native)
Store billingGoogle/Apple's mandatory payment system for digital goods sold inside store apps, with a 15–30% cut
White-label platformA service that generates a branded app/site for a business (e.g. Flipdish for takeaways)
Marketplace appA shared platform aggregating many vendors (Just Eat, Uber Eats, Deliveroo)
WebViewAn embeddable browser component an installed app can use to show web content inside itself; the basis of hybrid wrappers

Check your understanding

  • Why is an app installed from a store not necessarily a platform-native app?
  • What can a store do to a native app that nobody can do to your website?
  • Why does a fast food chain pay roughly 3% on an in-app order while a $69 digital book sold through store billing would lose 15–30%?
  • Where does a TWA sit on the spectrum between a plain web app and a fully native app?
  • Figma and McDonald's both care about mobile users; why does Figma's core product stay in the browser while McDonald's pushes regulars into a native app?

Reviewed 5 August 2026.