PWAs were hailed a few years ago as "the future" — a way to build one thing that works as both a website and an app. In 2026 the picture is more nuanced. Apple has kept PWA support limited, App Stores have tightened rules, and pure native apps have gotten new momentum. But PWAs aren't dead — they've just become a specific tool for specific problems.
What a PWA actually is
A PWA is a website with three additional features:
- Installability — users can "add to home screen" and get an app icon
- Offline support — via Service Worker that caches resources
- Push notifications — alerts even when the app isn't open
Plus: runs full-screen without browser UI, has a splash screen, and can integrate with platform features (camera, GPS, file system) — but here differences between iOS and Android persist.
Where PWAs win
1. You have a website, not an app idea
If you fundamentally have content that users consume (articles, dashboards, e-commerce) and don't need deep platform integration, PWA is often faster and cheaper than native. You build one thing, not three.
2. You want to avoid App Store friction
App Store and Google Play take 15–30 % of in-app purchases. If you sell something (subscription, in-app purchases), a PWA can save you significant amounts. Spotify, Netflix and others push users to web payment for exactly this reason.
3. You don't want to maintain three codebases
Web + iOS + Android = three versions that need to stay in sync. A PWA is one codebase. For SMBs without large teams that matters.
4. International users on low-end devices
In markets where users have 5-year-old Android phones with 32 GB of storage, a 5 MB PWA is a much better experience than a 150 MB native app they don't have room for. Twitter Lite, Pinterest Lite and Uber Lite exist exactly for this reason.
Where PWAs lose
1. iOS limitations
Apple has been clear they don't want PWAs to compete with the App Store. Push notifications on iOS have been limited (though they improved in iOS 16.4+), background processing is non-existent, and some Web APIs (Bluetooth, NFC) aren't supported.
If your main users are on iOS and you need real app functionality, PWA is a compromise.
2. App Store exposure
Users discover apps in App Store and Google Play. A PWA doesn't have that exposure — it has to be marketed like a website. If App Store discovery is part of your go-to-market strategy, you need native.
3. Complex native features
AR, complex camera processing, integration with HealthKit / Google Fit, CarPlay, deep Bluetooth use — all of this is limited or impossible in a PWA.
4. "Feels like an app"
A perfectly built PWA gets 80 % of the way to the feel of a native app. But the last 20 % — micro-interactions, native gesture response, transitions — is hard to replicate. For premium brands where feel is part of the brand, this can be a deal-breaker.
Technical requirements for a good PWA
Service Worker — properly configured
Most PWA problems come from poorly configured Service Workers. Aggressive caching gives "stale" data, too little caching gives poor offline. Use Workbox or next-pwa to handle it properly.
Web App Manifest
File that tells the browser how the PWA should appear (icon, splash, theme colour, display mode). Has to be correct for "add to home screen" to work.
HTTPS
Required. No exceptions.
Performance
PWAs need to perform like apps, not like websites. Lighthouse score of 90+ on both performance, accessibility and best practices is the minimum.
Tools for building PWAs
- Next.js + next-pwa — our favourite for new projects. Server-side rendering, automatic code splitting, and PWA functionality with minimal config
- Vite + PWA Plugin — for SPAs where SSR isn't needed
- Workbox — Google's library for Service Worker patterns, works with any stack
Actual use cases where PWAs work best
- News publications — users consume content, want offline access
- SaaS dashboards — pro users, more often on desktop, want "always on" feel
- B2B tools — internal use, App Store discovery not relevant
- E-commerce for mid-sized stores — alternative to App Store friction
- Frequent updates — apps that change often and need immediate deployment
What we recommend
- Building something that's primarily content + light interaction? → PWA is likely the right choice
- Need deep platform integration, AR, or App Store discovery? → Native or cross-platform native
- Unsure? → Build PWA first, and add a native wrapper with Capacitor if you later find you need App Store presence
At Inovix we've built both PWAs and native apps. If you're weighing which fits your project, we can give an honest assessment in 20 minutes.
