Does my small business actually need a mobile app?
Most SMB owners who ask this question do not need a native app. They need a better website, a PWA, or an off-the-shelf booking tool. The six tests below tell you which group you are in.
10 min read
Scope my app →The average SMB app costs USD 18,000–32,000 to build and achieves under 12% 30-day retention. Most of them are replaced by a USD 8,000 PWA eighteen months later — after the owner has paid App Store review fees, watched installs stall, and missed two iOS major releases. Start with the right question.
The expensive default
The conversation usually starts like this: "I want to launch an app for the business." A developer hears USD 60k, an offshore shop hears USD 25k, a no-code consultant hears USD 6k. None of them ask whether you should build at all.
We will. Below are six questions. If you answer yes to at least four, a real app is worth the cost. Fewer than four, look at lighter options first.
Test 1: Frequency of use
Will your customer open the app at least once a week? Banking apps pass. Food-delivery apps pass. Annual booking apps fail — they get downloaded once for the booking and uninstalled the day after.
Median 30-day retention for SMB-published apps is around 12%. For apps that solve a weekly workflow, retention can hit 35–55%. The difference is not marketing; it is fit between the app and the user's rhythm.
Test 2: Hardware needs
Does the experience genuinely require things only a native app can do?
- Camera with custom processing (barcode scanning beyond the basic web API, AR overlays, real-time filters).
- Bluetooth peripherals — payment terminals, fitness sensors, smart home gear.
- NFC for tap-to-pay or tap-to-access.
- Background location tracking (delivery driver, field service dispatch).
- HealthKit / Google Fit integration with consented data flows.
- Apple Pay / Google Pay in-store via the native APIs.
If none of these apply, a PWA can probably do what you need at half the cost.
Test 3: Distribution leverage
Will being listed in the App Store and Play Store materially help you get discovered?
For some categories — productivity utilities, games, indie SaaS — the stores are a real channel. App Store Optimization (ASO) drives meaningful download volume.
For most SMBs (a local service, a B2B tool, a single-brand DTC), the App Store is not a channel. Almost nobody downloads "Joe's HVAC" from the store. They install it from a link your business sent them after the first interaction. A PWA does that with one tap and no review process.
Test 4: Offline necessity
Does the experience need to work without internet?
Field service tools that operate in basements with no signal, in-store retail apps for sales floors with patchy WiFi, transit apps for the subway — yes. They benefit from native's persistent local storage and richer offline UI patterns.
Customer portals, booking flows, dashboards, e-commerce browsing — no. They can be served by a PWA with a service worker, which handles intermittent connectivity well enough for the use case.
Test 5: Push notification value
Will push notifications drive measurable behavior?
Push has matured. iOS now allows web push from PWAs (Safari 16.4+). Android has supported it for years. For most SMB use cases — order status, appointment reminders, low-stock alerts to staff — a PWA with web push gets you 80% of native push value at 40% of the cost.
Native push wins when notifications need rich media (images, video in the notification UI), interactive actions (snooze, reply), or critical-alert privileges (rare and Apple-gatekeeper-controlled).
Test 6: Acquisition cost reality
The hidden cost of an app is not the build. It is the cost of getting people to install it.
Average install cost via paid ads in 2026 for SMB apps: USD 4–18 depending on category. Then a 65–80% drop-off between install and first meaningful action. Effective cost per active user is often USD 20–60.
Compare that to a website lead at USD 8–30. If your unit economics cannot absorb the install funnel, the app cannot exist as a standalone product no matter how well it is built.
PWA first: the option most SMBs overlook
Progressive Web Apps are the most underrated answer to the "do we need an app" question. For the majority of SMB use cases — customer portals, booking flows, loyalty cards, member dashboards — a PWA delivers the experience at a fraction of the native cost.
A PWA:
- Installs from a URL with one tap — no App Store review, no download wait, no friction.
- Lives on the home screen with an icon and runs full-screen, indistinguishable from a native app for most users.
- Works offline via a service worker.
- Supports push notifications on iOS 16.4+ and all modern Android.
- Updates instantly — no app-store-submission delay between bug fix and customer.
- Uses the same Next.js / React stack as a modern website — your team can maintain it.
Cost comparison: PWA vs React Native vs native
| Build type | Build cost (MVP, 6–10 screens) | Annual maintenance | App Store review? | Hardware access |
|---|---|---|---|---|
| PWA (Next.js) | USD 6,000–12,000 | USD 1,200–3,600 | No | Camera, GPS, push (limited) |
| React Native (Expo) | USD 18,000–32,000 | USD 6,000–12,000 | Yes (both stores) | Full iOS + Android APIs |
| Native iOS + Android | USD 50,000–80,000+ | USD 14,000–24,000 | Yes (separate submissions) | Full platform-specific APIs |
When you actually need native
Four specific criteria justify native over PWA. All four should be evaluated against your actual use case, not your aspirational roadmap.
- Background processing required. GPS polling while the app is backgrounded (delivery driver tracking, field-service dispatch), audio processing in the background (recording, streaming), Bluetooth LE device polling. PWA service workers can handle limited background sync but cannot do persistent sensor polling.
- Hardware integration. Bluetooth LE peripherals (payment terminals, fitness sensors, IoT devices), NFC tap-to-access or tap-to-pay, HealthKit / Google Fit consented data flows with write access, camera with real-time processing (AR, live barcode scanning at high throughput).
- Critical push notifications.Rich media in the notification UI (images, video), interactive notification actions (reply, snooze without opening the app), Apple's Critical Alerts that bypass Do Not Disturb for medical or emergency use cases.
- Offline-first with complex local data sync. Field service apps where workers fill forms in basements or rural areas and sync on return. Retail POS that must function during outages. These require robust local database handling (SQLite via Expo, Core Data, Room) that PWA IndexedDB cannot replicate reliably.
The decision summary
| Tests passed | Recommendation | Typical cost |
|---|---|---|
| 0–1 | Skip the app. Improve the website or buy off-the-shelf software. | USD 0–8,000 |
| 2–3 | Build a PWA. No App Store dependency, 40–60% cheaper than native. | USD 6,000–12,000 |
| 4–5 | React Native via Expo. Single codebase, ships to both stores. | USD 18,000–32,000 |
| 6 | Native Swift + Kotlin for platform-specific surfaces, shared logic layer. | USD 50,000–80,000+ |
When IF Digital recommends building the real thing
We have shipped native apps for: a multi-location dental practice (HealthKit integration, scheduling, intake forms with offline signature capture), a regional grocery chain (in-store barcode scanner + loyalty + Apple Pay), and a field-service operator (offline work-order forms, GPS-based dispatch, photo capture with annotation).
All three passed four-plus tests. All three are still in active use 18+ months later. The apps that fail are the ones built because "we should have an app" without a single test passing.
Bottom line
Most SMBs asking "do we need an app?" need a PWA. The PWA delivers 80–90% of what a native app delivers for the typical SMB workflow at 40–60% of the cost, with no App Store dependency, instant updates, and the same team maintaining it as your website.
Native makes sense when background processing, hardware peripherals, critical push, or complex offline sync are real requirements — not when they are on a theoretical roadmap. If you are unsure which tests apply, bring us the specific workflow you want to digitize and we will tell you the right build for it.
Run the six tests against your business. If you land at 2–3 passes, send us your workflow and we will scope a PWA. If you land at 4+, send us the workflow and we will scope a React Native MVP. If you land at 0–1, we will tell you straight that you do not need an app and probably need a better website instead.
Frequently asked questions
- What's the cheapest way to get an app-like experience without building a real app?
- A Progressive Web App (PWA). Built on the same Next.js or React stack as a modern website, installable to a phone home screen from the URL, works offline with a service worker, and can send push notifications on Android and iOS 16.4+. Cost: USD 6,000–12,000 for a focused SMB PWA. That is roughly 40–60% less than a React Native build for the same feature set — and it skips App Store gatekeeping and review delays entirely.
- How much does a real native mobile app cost for an SMB?
- A focused MVP using React Native via Expo, single codebase for iOS and Android, runs USD 18,000–32,000 for 6–10 screens. Native Swift + Kotlin (separate codebases) runs USD 50,000–80,000+ for the same scope. Annual maintenance averages USD 8,000–18,000 to keep up with iOS/Android OS releases and SDK deprecations. Budget for one major SDK-breaking change per platform per year.
- How long does the App Store approval process take in 2026?
- Apple averages 1.5–4 days for a first review. About 35% of first submissions are rejected for fixable reasons: missing privacy labels, demo credentials not provided, in-app purchase rules violations. Resubmission turnaround is similar. Plan two weeks from build-complete to store-live as a healthy buffer. Google Play is faster: 1–3 days for first review, ~20% first-rejection rate.
- Will an app actually be used, or will it sit on the user's phone unopened?
- Median 30-day retention for SMB-published utility apps is under 12%. The exceptions are apps that solve a recurring workflow users complete at least weekly — loyalty check-ins, booking, ordering, status checks, communication with the business. Before building, answer: does your customer do this specific thing on their phone more than once a week? If no, the app will sit unused regardless of how well it is built.
- When do you actually need a native app instead of a PWA?
- Four specific criteria: (1) background processing — GPS tracking, audio processing, Bluetooth polling while the app is not on screen; (2) hardware integration — Bluetooth LE peripherals, NFC tap-to-pay, HealthKit/Google Fit consented data flows; (3) critical push notifications — rich media, interactive actions, Apple critical alerts that bypass Do Not Disturb; (4) offline-first with complex local data sync — field service apps, retail POS, transit. If none of these apply, a PWA covers the use case at lower cost.
- Can a PWA replace the App Store distribution channel?
- For most SMBs, yes. Almost nobody finds a local service business's app by browsing the App Store. They install it via a link from a booking confirmation email, a QR code on a receipt, or an in-store prompt. A PWA installs in one tap from any of those entry points with no App Store friction. The App Store is a meaningful distribution channel for consumer utility apps, games, and indie SaaS — not for most local or B2B SMB use cases.