How SEM Nexus Picks Between Flutter, React Native, Angular, and Native — In Two Weeks

Most mobile-dev agencies sell you a stack before they understand your project. SEM Nexus runs the stack analysis as part of every two-week discovery — and at the end of week 2, you have a written technical recommendation that names the stack, names the reasoning, and is signed by the senior engineer who'll write the architecture.

This is the exact process we run inside discovery, the five questions that determine the answer, and the reason the SEM Nexus stack recommendation has been right on every single build we've shipped in the last two years. Not because we're lucky. Because the process forces the right answer.

The five questions that determine the stack

We work through these in order. The answer to each narrows the choice.

1. What's the hard part of the project?

Every mobile build has one or two things that are 10x harder than the rest. Audio engines. Offline-first sync. Real-time location. BLE integration. Custom UI rendering. On-device ML. Regulatory work.

If the hard part has a clearly best stack, the decision is essentially made. Cerebyte needed an audio engine — Flutter had the strongest audio ecosystem in 2024. Decision made before we got to question 2.

If the hard part doesn't have a clearly best stack — most builds — we go to question 2.

2. What stack does your team or your future team know?

Stack picks are often more about the team than the project. My Home Delivery's founder had a React-shop team running a Laravel back-end. Going Flutter would have meant introducing Dart to a team already running PHP + JavaScript — real onboarding cost. React Native picked itself.

If you're a solo founder planning to hire mobile engineers in 12–18 months, we look at the hiring market in your geography. React Native has the deepest pool. Flutter has a smaller but rapidly growing pool. Native iOS/Android has the most senior engineers but pays more.

3. How much custom UI does the app have?

If the app's UI is mostly lists, forms, tabs, and modals — the standard mobile-app shapes — Flutter and React Native are roughly equivalent. Either works.

If the app has highly custom UI — drawing tools, video editors, game-adjacent experiences, animation-heavy interfaces — Flutter's "draws its own pixels" model wins. Impeller renders consistently across platforms in a way RN's bridge cannot.

If the app is mostly text/images thinly wrapping content — a CMS-driven app, a news app, a thin content portal — Angular + Ionic is genuinely the cheapest and shipping-est option, and our 360 Medical Consulting and MyPace builds proved it.

4. Does the app's value depend on a native-first feature?

If the central pitch involves Live Activities, Dynamic Island, ARKit, on-device ML, BLE with weird hardware vendors, Health/HealthKit integration, App Clips/Instant Apps — go native. The cross-platform wrappers exist but lag the native SDKs by months, and you'll fight the abstraction more than you'll enjoy the shared codebase.

We've shipped native iOS-first builds when the value depended on it. We don't push native when it isn't justified — and most B2B/B2C apps don't need it.

5. How fast does your project need to ship?

For a typical 4-month SEM Nexus build:

• Native adds 30–50% to the timeline (two codebases, one team).
• React Native and Flutter are roughly equivalent.
• Angular + Ionic is the fastest if the team has web skills.

If you need to be in the App Store in 16 weeks, cross-platform is the realistic call for most B2B/B2C projects.

Want this analysis applied to your project, written down, with the reasoning your team can challenge? Book a two-week discovery and we'll deliver the technical recommendation at the end.

The written deliverable

At the end of week 2, every SEM Nexus discovery produces a one-page technical recommendation. The structure is consistent across every project:

1. Recommended stack — Flutter / React Native / Angular + Ionic / native iOS+Android
2. Reasoning — which of the five questions drove the answer, with specific evidence
3. Hard part identified by name — the technical risk that determines whether the project ships on time
4. Key integrations — payments, auth, push, analytics, your back-end API, etc.
5. Architectural risks named in writing — what could go wrong, what we'd do if it does
6. Stack-specific tooling — the libraries, frameworks, and patterns we'll use

The document is one page. Founders can read it in 5 minutes. Their CTO advisors can argue with it in 30. We've had advisors push back, propose alternatives, and force us to defend or revise the recommendation. That argument is the value. A recommendation that nobody can argue with is a recommendation nobody understood.

Three real recommendations and what they produced

Cerebyte: Flutter + Node.js.

Hard part: real-time audio engine playing back EEG-derived signals on iOS and Android with low latency. The recommendation cited Flutter's just_audio + Dart FFI as the cleanest path to writing the engine once. Native was considered and rejected because the duplication risk was too high. React Native was rejected because the audio ecosystem in 2024 didn't support the real-time mixing the project needed.

Outcome: shipped in 14 weeks. Audio latency within 5ms of native. The stack choice held.

My Home Delivery: React Native + Laravel.

Hard part: Stripe Connect multi-party payments + real-time driver location tracking. The recommendation cited the founder's existing React team as the deciding factor — the engineering team could read and contribute to both the web admin (React) and the mobile app (React Native) without learning a new language.

Outcome: shipped in 16 weeks. Six months later, the founder hired their first internal mobile engineer in a week — React Native hires are everywhere.

360 Medical Consulting: Angular + Ionic.

Hard part: HIPAA compliance + EHR integration. The recommendation cited the practice's existing Angular web admin and the maturity of the Ionic ecosystem for content-and-forms-heavy apps. Native was overkill; React Native and Flutter would have introduced a third language to a team already running Angular and PHP.

Outcome: shipped HIPAA-compliant patient portal in 20 weeks (extra 4 for compliance work).

What this process catches that ad-hoc stack choice misses

The five questions are designed to surface tradeoffs you didn't know to ask about. Founders often arrive convinced they want Flutter ("the modern choice") or React Native ("our team knows JS"). Sometimes they're right. About 30% of the time, the discovery analysis points at a different stack.

The honest answer is rarely the answer the founder walked in with. That's the value of running the analysis — the agency that just ships you whatever you walked in with is selling you their specialty, not the right answer.

If you're scoping a build and you'd like the analysis applied to your project, SEM Nexus's two-week discovery includes the stack recommendation as a deliverable. The reasoning is on paper. Your team can challenge it. We'd rather defend a hard recommendation than ship the wrong stack quietly.