The bottom line
On September 10, 2026, Shopify announced it is moving its mobile apps from React Native back to separate native codebases written in Swift (iOS) and Kotlin (Android), reversing a strategy it had championed since 2020 (Shopify Engineering). The company's own explanation is not that React Native failed — it explicitly calls the framework "excellent" and says it delivered on its original promises (Shopify Engineering). Instead, Shopify says AI coding agents have gotten good enough that building the same feature twice, once in Swift and once in Kotlin, no longer costs what it used to. This is a single company's self-reported account of its own engineering decision, not an independent study, so the concrete numbers (timelines, team sizes) should be read as Shopify's own figures, not third-party-verified ones.
What happened and how Shopify tells it
Shopify's history with mobile frameworks has looped back on itself. In 2020, the company went "full steam ahead" on React Native for three stated reasons: to stop building the same feature twice, to let developers work across the whole stack, and to spend less time chasing parity between iOS and Android versions of the same app (Shopify Engineering, via Hacker News timeline). As recently as January 2025, Shopify wrote publicly that React Native's limitations "simply don't exist anymore" and encouraged other teams to revisit it (HN summary of that post). In September 2025, Shopify said it had successfully migrated its two largest apps — Shopify Mobile and Shopify Point of Sale — to React Native's "New Architecture," a significant technical investment (HN summary).
A year later, that investment path was abandoned. Shopify says the trigger was practical: it was about to commit to another round of New Architecture work when it decided to first test whether coding agents could build directly in native tools — SwiftUI on iOS and Jetpack Compose on Android — while keeping the two platforms' behavior aligned (Shopify Engineering).
What they found
Shopify describes a staged process, not a single leap. First, a proof of concept: one engineer spent one week using coding agents to reproduce as much of the existing React Native "Shop" app as possible in native SwiftUI, using the existing app as a reference. The result wasn't production-ready, but Shopify says it was convincing enough to justify a full migration, and that agents worked especially well when given an existing implementation to copy from — porting features, scaffolding screens, wiring up data, and refining layouts from visual feedback (Shopify Engineering).
From there, a core team of six engineers built the native foundations and main user flows, with feature teams joining partway through. Shopify and a report on the story both put the total timeline for the Shop app's rebuild — from proof of concept to a fully rebuilt app published in the app stores — at twelve weeks (Shopify Engineering; The New Stack). As a consequence of the shift, Shopify is also winding down some of its own open-source React Native contributions: two libraries, react-native-skia and flash-list, are being handed to new maintainers, while a third, restyle, will be archived at the end of 2026 because of its smaller user base (Simon Willison's summary).
What this does not show
This is one company describing its own project, published on its own engineering blog, which also functions as recruiting and PR material — that framing matters when weighing how the numbers are presented. There's no independent audit of the "one week" proof of concept, the "six engineers," or the "twelve weeks" figure, and no comparison group showing what a from-scratch native rebuild would have cost without AI assistance, or what continuing with React Native's New Architecture would have cost instead. The post also doesn't measure user-facing outcomes — it argues about engineering cost, not app performance or customer satisfaction. Separately, discussion on Hacker News raised a tangential but unverified point: one commenter estimated Shopify employs around 3,000 engineers as of 2026 and contrasted that with Chrome's roughly 60 engineers at its 2008 launch and GTA V's 150-person credited engineering team, arguing Shopify's apps are overbuilt (Hacker News thread). Another commenter pushed back, noting this compares an entire company's headcount to a single product's launch team, and that early Chrome was missing basic features like PDF export and accessibility support that later required hundreds of engineers to build (Hacker News thread). That side debate is about Shopify's overall engineering scale, not about whether the React Native decision itself was sound, and neither side's numbers were verified.
What it means, if anything
Shopify is a large, well-resourced company that could reassign a dedicated team of six engineers for three months to a native rebuild — a resource pattern many smaller teams don't have. Its stated reasoning is narrow and specific: coding agents changed the relative cost of building twice, not that cross-platform frameworks are technically worse. Whether this generalizes to other companies, especially smaller ones without spare engineering capacity or teams that value having non-mobile-specialist developers contribute to mobile code (one of React Native's original selling points for Shopify), is not something this case study can answer. For now, this is one high-profile, self-reported example of a changed cost calculation — suggestive of a trend AI coding tools might be enabling, but not evidence that the trend is widespread or that it will hold for teams with different constraints.
Further reading
- Native is now the future of mobile at Shopify — Shopify Engineering
- Migrating Shop app from React Native to native — Shopify Engineering
- Native is now the future of mobile at Shopify — Simon Willison's Weblog
- Shopify spent years on React Native — then rebuilt everything in 12 weeks — The New Stack
- Hacker News discussion