You think launching a mobile app requires two native teams, a six-figure budget, and 18 months of development? That playbook is obsolete. The mobile market has shifted without most technical decision-makers noticing, and traditional agencies have every incentive to keep things murky.
I run mobile projects from Vietnam for French clients, and what I see on the ground contradicts most of the advice you'll find in IT consultancies' sales decks. Native apps are losing ground to cross-platform. AI lets a senior dev deliver in three weeks what used to take three months. Vibe coding produces impressive prototypes that blow up in production.
- 📉 Native in decline · Swift and Kotlin are losing share to React Native and Flutter.
- ⚡ Cross-platform dominant · a single codebase covers iOS, Android, and web.
- ⚠️ Vibe coding is misleading · prototyping is not shipping a maintainable product.
- 🎯 The winning formula · a senior offshore team augmented by AI, not more devs.
Native development is dying (and nobody is telling you)
The Stack Overflow surveys and GitHub trends are clear: Swift and Kotlin, still dominant five years ago, are being overtaken by JavaScript, TypeScript, and cross-platform frameworks. This is not a weak signal. It is a structural shift.
Why are companies abandoning native?
The reason is economic before it is technical. Maintaining two parallel codebases (iOS + Android) with two specialized teams costs double for what is often an identical result from the user's perspective. Job postings reflect this pivot: recruiters are looking for React developers who can do mobile, not the other way around.
Even on the user side, behavior is changing. People no longer download new apps the way they used to. They stick to their core apps and handle the rest through the mobile web. Experienced developer CodeHead sums it up well: "More companies are building responsive web apps than native mobile apps from scratch."
Native still makes sense for apps with heavy hardware interaction (real-time camera, AR, 3D games). For a SaaS, a marketplace, or an internal business tool? It is a waste of budget.
When does native still make sense?
Three specific cases: apps that leverage specialized sensors (health, IoT), those requiring extreme graphical performance, and those whose business model relies entirely on the tactile experience. For everything else, cross-platform delivers 95% of the result at 50% of the cost.
Cross-platform and PWA: the new technical standard
React Native, Flutter, and Progressive Web Apps have redefined expectations. A senior developer proficient in React Native can deliver a functional app on iOS and Android with a single codebase, a single deployment pipeline, and a single team.
How did React Native conquer the market?
Instagram, Shopify, Discord: these apps run on React Native. The framework has matured to the point where it covers 90% of professional mobile use cases. The argument that "native performance is superior" holds less and less weight given recent optimizations and the power of today's devices.
AristiDevs, an experienced mobile developer, puts it bluntly: "The best language is the one that makes you money." That statement hits especially hard when you compare the cost of a native team (2 iOS devs + 2 Android devs + coordination) to a cross-platform team (2-3 React Native devs shipping everywhere).
On our projects at GoLive Software, the React Native vs Flutter decision depends on the client's ecosystem. React Native if the internal team already knows React (which is the case for most SaaS startups). Flutter if the project starts from scratch and UI performance is critical.
What is the real advantage of PWAs?
PWAs add another layer: installable, capable of working offline, and accessible without going through the stores. For an internal tool, a client portal, or a low-retention app, the PWA eliminates download friction and Apple/Google's 30% commission.
| Approach | Relative cost | Time-to-market | Coverage | Trend |
|---|---|---|---|---|
| Native (iOS + Android) | 100% | 6-12 months | 2 platforms | ↓ declining adoption |
| React Native / Flutter | 50-60% | 3-6 months | iOS + Android + Web | ↑ +40% job postings |
| PWA | 30-40% | 2-4 months | All browsers | ↑ B2B adoption |
| Native + PWA hybrid | 70-80% | 4-8 months | Everything | → niche cases |
SOURCE: Stack Overflow Developer Survey 2025 + GitHub Trends · Updated 05/2026
Vibe coding for mobile: miracle or mirage?
A recent Reddit post made the rounds in the dev community: a developer claims to have "vibe coded" over 12 mobile apps to reach 500,000 downloads and 100,000 monthly active users. His stack: React Native Expo + Firebase/Supabase, driven by Claude Opus. Impressive on paper.
Why does vibe coding work for some people?
This developer already had solid technical foundations (Udemy courses, YouTube experience). He no longer writes a single line of code himself, but he understands the architecture. That is where his success comes from: he knows what to ask the AI, he knows how to validate the output, and he knows when the generated code is shaky.
Another Reddit testimonial confirms the pattern: a dev hitting $1,000/month with 14 apps (health trackers, habit trackers). His key? Naming the app exactly what people search for in the store ("Sober Tracker," not a creative name), switching to subscriptions instead of one-time purchases, and iterating fast.
These successes are not reproducible by a non-engineer. I see it on every project: producing code with AI does not mean knowing how to build a product. Architecture, security, edge case handling, data migrations, store compliance: all of that requires a real engineer behind the prompt.
How does vibe coding become dangerous without oversight?
The problem never shows up at the prototype stage. It shows up at scale, at the first critical production bug, at the first forced update from Apple. A product generated quickly without technical oversight can cost ten times more to fix than to build correctly from the start.
Jeremy Pitault, who built a language-learning app generating $10,000/month in MRR with 130,000 users, makes one key point: "If it's your first app and you've never managed to build one that works, go with something you know." Vibe coding does not eliminate this requirement for domain mastery.
"Producing code with AI doesn't mean knowing how to build a real product. Architecture, security, maintenance: that's still an engineer's job."
, Vincent Roye, May 2026
The hidden costs your agency won't mention
Initial development often represents less than 40% of a mobile app's total cost over three years. Nobody talks about the rest when it's time to sign.
What are the real post-launch expenses?
Bug fixes and OS updates, store compliance (Apple guidelines that change every six months), security patches, backend infrastructure, monitoring, and above all: feature evolution. An app that stops evolving dies in the stores.
Attorney Nicola Ferrante highlights an often-overlooked aspect: intellectual property ownership. A mobile development contract must clearly specify who owns the code, assets, and data. Without that clause, you can find yourself held hostage by your agency when it's time to switch teams.
A Reddit comment about the Home Depot app sums up the paradox of large companies: "I always wonder how these huge corps manage to make such bloated steaming piles of garbage. Meanwhile one talented person cranks out something better." The sluggishness of corporate apps rarely comes from a lack of budget. It comes from layers of tracking, data harvesting, and accumulated technical debt.
How to structure a realistic mobile budget?
My advice after dozens of delivered projects: budget 60% of the initial cost for annual maintenance. If your agency quotes you a flat fee of €80,000 for V1 without mentioning the €40,000/year that follows, they are selling you a mirage.
The startups that succeed don't burn their budget on a perfect V1. They launch a cross-platform MVP in 8-12 weeks, validate product-market fit, then iterate with a lean but senior team. That is exactly the model I advocate: a small, structured offshore team that ships fast, led by someone who understands the product.
The equation that changes everything: offshore + AI + seniority
The "onshore vs offshore" debate has become obsolete in its traditional form. The real question in 2026: does your team know how to use AI to multiply its velocity?
Why does AI strengthen the advantage of structured offshore?
A senior Vietnamese developer who masters Claude Code or Cursor now delivers what a team of three junior Parisian developers used to deliver two years ago. The cost? One third. The quality? Higher, because technical experience guides the architectural decisions that AI cannot make on its own.
I see it on my projects every week: AI does not replace good developers, it amplifies their output. A skilled, well-equipped dev ships in three weeks a sprint that used to take six. The difference is no longer about the number of developers, but about their ability to use AI intelligently.
According to Statista, the mobile app market exceeds $500 billion in annual revenue. The companies capturing a share of that market are the ones that ship fast, iterate fast, and maintain clean code. Not the ones with the biggest budget.
Should you keep mobile development in-house or outsource it?
If you are not a company whose core product is the app itself, outsource. Structured offshore development gives you access to seniors who have already shipped dozens of apps, without the constraints of hiring in the tight French market.
The future belongs to augmented developers, not replaced developers. And you will find those augmented developers more easily in Vietnam than in Paris, at a third of the daily rate, with the same stack and the same AI tools.
Frequently asked questions
How much does mobile app development cost in 2026?
A cross-platform MVP (React Native or Flutter) with a lightweight backend ranges from €25,000 to €60,000 depending on complexity. With structured offshore, divide by two to three. The total cost over three years (maintenance included) typically reaches 2 to 3 times the initial budget.
React Native or Flutter: which one should I choose for my project?
React Native if your team already knows React or JavaScript. Flutter if you are starting from scratch and UI performance is critical. Both cover 90% of use cases. The choice should be based on your existing ecosystem, not theoretical benchmarks.
Can vibe coding replace a development team?
For a prototype or a simple app (tracker, quiz, utility), yes. For a product with business logic, payments, compliance, and thousands of users, no. Vibe coding without technical oversight creates debt that costs dearly at the first incident.
Is it still worth developing in native Swift or Kotlin?
Only if your app makes heavy use of the hardware (AR, health sensors, 3D games) or if your users expect a premium experience indistinguishable from a system app. For the 80% of business apps, SaaS, or marketplaces, cross-platform is the right call.
How do I choose an agency for a mobile app?
Verify that they offer cross-platform (not only native, which signals an outdated approach). Ask for references on apps that have been in production for over a year. Make sure the contract covers code ownership and plan a maintenance budget from day one.
Vidéos YouTube
- Here's how I built an app to $100K ARR · Jeremy Pitault
- We Need To Talk About Mobile Dev... · CodeHead
- Developing Mobile Apps in 2025 - Which Programming Language Should You Use? · Programación by AristiDevs
- Professionelle App Programmieren: 8 Dinge, die du können solltest · Programmieren lernen
- Contratto legale per sviluppo di applicazione mobile · Avv. Nicola Ferrante
