Route memo
A short written comparison of the routes that could fit, scored against your team, your users’ phones, your design and the hardware the app touches. You receive: the memo, with our first and second choice and the reasons.
Cross-platform app development
We recommend the route that fits your team and your app in writing, then build it in accounts and a repository that belong to you.
Get a written route recommendation and scopeWhat cross-platform means
Cross platform app development is building one app for iPhone and Android without writing every screen twice. It comes in several forms, from sharing only the logic to sharing the whole app, and each form saves different work and costs different things.
React Native, Flutter, Kotlin Multiplatform, a web app on the home screen and two native apps are all ways to reach both stores. They differ in who can maintain the result and how close it sits to the phone.
Business rules, data fetching and many screens move between platforms well. Bluetooth, background work, payments sheets and permission prompts behave differently on each phone, and that work is not shared.
Before any build starts you get a route memo that names the option we would pick, the one we would pick second, and the facts about your team and app that decided it.
Choosing the approach
Both are cross-platform. They differ in how much of the app is written once, and so in what a change costs later.
| Shared logicKotlin Multiplatform with native screens | Shared screensReact Native, Flutter or Compose Multiplatform | |
|---|---|---|
| What is written once | Business rules and data handling, while each platform keeps its own screens | Screens, navigation and business rules |
| Who builds it | A shared-code engineer plus an iPhone developer and an Android developer for the screens | One team working in one framework, with a platform specialist on call for native pieces |
| How the app looks | Each platform’s own controls and conventions by default, with two sets of screens to keep in step | Consistent on both phones, with room to follow each platform’s conventions where it matters |
| Cost of a new feature | Shared part built once, two screens built separately | One build, then testing on both platforms |
| A good fit when | The company already has native developers and wants to share rules without changing how screens are built | Budget is tight, the design is the same on both phones and one team will own the app |
If the product is a camera pipeline, augmented reality, a game or a Bluetooth device with timing demands, two native apps are the safer route, and we would say so in the memo. If your users are all on one platform, one native app costs less than either.
The platforms and the frameworks
Each fact below names its source and the day we read it.
Google’s Android developer pages call it stable and production-ready, and say you choose what to share, from core business logic to the whole app. Sharing logic does not require sharing screens, which keeps native interfaces on the table.
The Compose Multiplatform 1.8.0 announcement says all major APIs are stable. The announcement cites the Respawn app sharing 96% of its code with Android. That is one app’s figure in the vendor’s own post, so read it as an upper bound.
Apple’s WebKit team announced Web Push for Home Screen web apps with iOS and iPadOS 16.4 beta 1 on February 16, 2023. A permission request has to follow a direct tap, such as a Subscribe button. A web app can send alerts, but it reaches people through a Home Screen install, with no store listing.
Apple’s requirement names an SDK for iOS 26 or later. Whatever framework wrote your app, its iPhone build is compiled in Xcode, so the framework version has to keep up with Apple’s periodic SDK minimum.
Android’s documentation says any app that uses NDK libraries, directly or through an SDK, has to be rebuilt for 16 KB devices. A cross-platform framework ships compiled native libraries inside the app, so in our reading the framework and plugin versions are part of the check.
Flutter’s documentation describes building a Flutter module into a host app written in Swift, Objective-C, Kotlin or Java. It lists limits for mobile: only the multi-engine approach is supported, and packing several Flutter libraries into one app is not. A working native app can gain Flutter screens without being rewritten first.
The scope
The work in six parts, each handed over as a document or a build.
A short written comparison of the routes that could fit, scored against your team, your users’ phones, your design and the hardware the app touches. You receive: the memo, with our first and second choice and the reasons.
A map of what is written once and what is written per platform, such as sign-in, payments, push setup and permissions. You receive: the boundary map, so nobody is surprised by platform work later.
A small throwaway build of the riskiest piece, such as Bluetooth, background location or a custom drawing, run on real phones before the scope is final. You receive: the spike result and the decision it led to.
Both apps built from one codebase in the framework we agreed, with platform code kept in marked files. You receive: the repository in your organization and a README a new developer can follow.
Test builds on older and newer phones from both platforms, with issues logged and fixed or disclosed in the release notes. You receive: the test log and the open-issue list.
A dated list of the changes coming from Apple, Google and the framework, such as SDK minimums and target levels, with who handles each. You receive: the plan and a monthly written scope option for the upkeep.
How it runs
Before the next stage begins, you approve the last one. The store accounts and the repository belong to you from day one.
Route decision
| Route | Fit | Reason |
|---|---|---|
| Flutter | Chosen | The slip map and tide view are drawn once, and one team can keep both phones in step |
| React Native | Second | Workable, but there is no JavaScript team or web app to share code with |
| Kotlin Multiplatform | Possible | Fits an in-house Android team, and would still mean two sets of screens |
| Web app on the home screen | Staff only | Fuel-dock forms fit a web app. Members expect a store listing and a Bluetooth gate unlock |
| Two native apps | Not now | Two builds for 1,400 member households, 68% of sign-ins on iPhone |
Our stack
The frameworks we compare, and the tools that build, test and watch the result.
Flutter is Google’s toolkit for building iPhone and Android apps from one Dart codebase.
We pick it when the design is custom and the same on both phones, and when no existing JavaScript team needs to share code.
Questions
Keep exploring
Business software built around how you work, in your own repository.
Swift and SwiftUI iPhone apps, built in your own Apple Developer account.
Kotlin and Jetpack Compose apps, published from your own Google Play account.
Engineers on your product, working in your GitHub, Jira and Slack.
A remote team building in your repository, with the code in your name.
A first version at a fixed price, with a list of what to cut.
Portals, dashboards and SaaS-style tools that run in the browser.
Software for larger organizations, built around the systems already in place.
One TypeScript codebase for iPhone and Android, in your own store and Expo accounts.
Start here
Tell us where the app starts today. We’ll send back the route we would choose and the runner-up, what is shared and what is per platform, and a fixed price for the first milestone.