Start typing to search this publication.
techcirkle
Open menu

Subscribe to techcirkle

Get new posts delivered straight to your inbox.

Native or Cross-Platform in 2026 — Deciding on Evidence

React Native and Flutter are production-grade. Native still wins in specific cases. Here is how to tell which case you are in, with actual criteria.

techcirkle
5 min read
Cover image for Native or Cross-Platform in 2026 — Deciding on Evidence

This argument used to be interesting. It mostly is not any more, which is progress.

React Native and Flutter are both production-grade and power applications with tens of millions of users. The performance objections that were legitimate in 2018 have largely been addressed. For the majority of commercial apps, cross-platform is the correct default.

What remains is a set of specific conditions under which that default flips. They are knowable before you write any code, which makes this a decision you can actually make on evidence rather than preference.

The savings, honestly stated

Cross-platform typically saves 30–40% against building and maintaining two native codebases. Not 50%, which is the number usually quoted.

The gap between the claimed and actual figure is platform-specific work that does not disappear:

  • Platform conventions — navigation patterns, back-button behaviour, share sheets, permission flows.

  • Store submission, signing, provisioning, and two distinct review processes.

  • Native modules for anything not covered by the ecosystem.

  • Testing across two platforms, which is the same work either way.

  • Divergent OS bugs that manifest on one platform only.

Roughly 60–70% of the code is genuinely shared. The rest is not, and pretending otherwise is where cross-platform estimates go wrong.

When native is the right answer

Heavy graphics or custom rendering. Games, image and video editing, AR, anything doing per-frame custom drawing. The bridge overhead is real at that level and the tooling is better natively.

Deep camera work. Not "take a photo" — that is fine cross-platform. Manual exposure control, custom capture pipelines, real-time frame processing, depth data. The abstractions leak badly here.

Complex background processing. Background location, continuous audio, health data sync, long-running tasks under aggressive OS power management. Both platforms handle this differently and both handle it badly; you want direct access to the platform APIs rather than a wrapper over them.

Day-one platform feature support. If your product depends on adopting new OS capabilities as they ship, cross-platform frameworks lag by months. For most products that is irrelevant; for a few it is the entire proposition.

Extreme binary size or startup constraints. Relevant mostly for emerging markets on low-end Android hardware, where a few megabytes and a second of cold start genuinely affect adoption.

When cross-platform is clearly right

Everything else, essentially. Content apps, commerce, booking, dashboards, social, most fintech front-ends, internal enterprise tools, marketplaces.

The argument gets stronger the smaller your team. Two native codebases require either engineers who work in both ecosystems — rare and expensive — or two separate teams, which means two sets of decisions, two release cadences, and divergent behaviour nobody intended. For a team under about eight engineers, that overhead is usually decisive on its own.

React Native or Flutter

If cross-platform is right, the choice between them is less consequential than it appears.

React Native suits you if your web stack is React and you want shared talent, shared patterns and a shared mental model. The ecosystem is enormous. The architecture rework of recent years addressed most of the historical bridge complaints.

Flutter suits you if you want maximum control over rendering and highly consistent visual output across platforms. It draws its own widgets rather than mapping to native ones, which is a genuine advantage for distinctive design systems and a mild disadvantage when you want to feel exactly like the platform.

Both are viable for essentially any commercial app. Pick based on your existing team, not on benchmarks — the difference in outcomes from team familiarity dwarfs the difference between the frameworks.

The signal worth watching

When you ask an agency this question, listen to the shape of the answer rather than the conclusion.

A good team asks what your app actually does before answering. They will want to know about background processing, camera requirements, whether you have a design system, what your web stack is, how large your team is, and what your maintenance plan looks like. Then they will give you a recommendation with the trade-off explained in terms of your specific product.

A team that recommends the same technology to every client is telling you about their hiring, not about your product. That is not automatically disqualifying — a Flutter specialist genuinely good at Flutter is a reasonable thing to hire — but you should know that is what you are getting, and you should be confident it fits.

One practical note on the decision

This choice is reversible at real but not catastrophic cost, for roughly the first six months.

After that, the cost rises sharply — not because of the code, but because of accumulated platform-specific workarounds, native modules, and institutional knowledge. Teams that switch after a year usually end up rewriting rather than porting.

So it is worth spending a week on the decision and not worth spending a month. Get the obvious disqualifiers right — background processing, camera depth, custom rendering — and move.

The full guide to evaluating UK app companies covers costs, archetypes, compliance and contract terms: read it here.

Frequently Asked Questions

How much does cross-platform development actually save?

Around 30–40% against building and maintaining two native codebases, not the 50% often quoted. Roughly 60–70% of code is genuinely shared; the remainder is platform conventions, store submission, native modules, per-platform testing and divergent OS bugs, none of which disappear.

When should I choose native over React Native or Flutter?

For heavy graphics or custom per-frame rendering, deep camera work such as manual exposure or real-time frame processing, complex background processing under aggressive power management, day-one support for new OS features, or extreme binary-size and startup constraints on low-end hardware.

Is React Native or Flutter better?

Neither, in general. React Native fits teams already working in React, offering shared talent and patterns. Flutter offers more rendering control and highly consistent visuals across platforms because it draws its own widgets. Choose based on your existing team's familiarity — that difference outweighs anything in the frameworks themselves.

Can I change my mind later?

For roughly the first six months, at real but manageable cost. After that it rises steeply, not because of the application code but because of accumulated platform-specific workarounds, native modules and institutional knowledge. Teams switching after a year generally rewrite rather than port.

What does it mean if an agency always recommends the same framework?

That they are describing their hiring rather than your product. It is not automatically disqualifying — a genuine Flutter specialist is a reasonable thing to hire — but you should recognise what you are buying and confirm independently that the framework fits your requirements, particularly around background processing and camera use.

Subscribe to techcirkle