'We want an app' is usually the wrong question. The right one: what behaviour are we enabling? For 80% of small businesses a PWA is cheaper, faster to ship and more effective than a native app. When to choose what — with numbers.
1. What a PWA actually is (no buzzwords)
A Progressive Web App is a website that behaves like an app: installable on the home screen, works offline, push notifications, native-like UI. One codebase for iOS, Android and web. No App Store needed — visitor taps 'Add to home screen'.
2. Cost comparison (realistic 2026 ranges)
- PWA (1 codebase): €8,000–25,000 build · €0–500/month maintenance
- Native (iOS + Android, 2 codebases): €40,000–120,000 build · €1,500–4,000/month
- Cross-platform (React Native/Flutter): €20,000–60,000 build · €800–2,000/month
3. When a PWA is enough (most small businesses)
- Order / book / reserve via your site
- Client portal with invoices, projects, tickets
- Webshop with checkout
- Loyalty program, discount codes, push notifications for offers
- Dashboard for internal workflows
4. When you actually need native
- Bluetooth/NFC/hardware integration that doesn't work via web on iOS
- Continuous background tracking (sports apps, GPS routes for hours)
- App Store as marketing channel (B2C product with search volume in the Store itself)
- Apple/Google Pay native integration in high-volume checkout flows
5. The hidden App Store costs
Native app means: Apple Developer account (€99/year), Google Play (€25 one-time), 15-30% commission on in-app payments, 3-7 day review cycles for every update, and broken app versions on major OS releases (twice a year). PWAs don't have that — push as often as you want.
6. The installation paradox
Many small business owners think 'an app feels more professional than a website.' But 70% of native apps that visitors download get deleted within 7 days. PWAs have no install friction: one tap, no app store, no 50MB download. For small business use that's an advantage, not a disadvantage.
7. How we usually advise
Start with a PWA. Measure for 6 months how often users install it, how often they return, what they miss. If a real hardware blocker or significant App Store demand emerges — then native. Building on gut feeling ('customers want an app') leads to expensive projects nobody uses.