Guide · one sourced statistic, zero invented ones

    iOS or Android: which is better to start your mobile app on?

    The old answer was a bet: pick wrong and pay for a rewrite. The 2026 answer is smaller and calmer — pick who you serve first, because architecture finally made the second platform an addition instead of a do-over. Here is the honest framework, with the one market number that is actually checkable.

    Start with the only number this article will assert: worldwide, Android runs roughly two-thirds of the world's phones and iOS roughly one-third (StatCounter, August 2026 — exact figures in the table below). If apps were elected by device count, the article would end here.

    They are not, and every founder who has looked at this decision knows why: install base and value base are different maps. iOS concentrates in the markets and demographics where paid apps and subscriptions historically perform — the industry's persistent rule of thumb, which you should verify against your market rather than take from us or anyone. Android owns the breadth: most of the world's users, most of its growth markets, and categories where reach beats revenue-per-user.

    So "which is better" has no general answer — but your answer usually falls out of four questions in about ten minutes.

    The four questions that decide it

    1. Where do your first hundred users actually live? Not your dream market — the people you can reach this quarter. Check StatCounter's per-country pages for the places you'll launch: platform share varies enormously by country, and your product launches in specific countries, not "worldwide." A community app for your city follows your city's split; a tool for US designers follows a very different one.

    2. How does the app make money? Paid up front, premium subscription, high-price niche → the rule of thumb points iOS first. Ad-supported, scale-driven, emerging-market audience → Android's install base is the asset. Selling to businesses? Then the buyer's device fleet decides, not consumer statistics.

    3. What do you own and use? Unromantic but real: you will test the app every day, and your daily phone should run it. Building iOS requires a Mac regardless of tool. If you carry an Android phone and a MacBook, you can genuinely go either way; if you've never held an Android device, shipping Android first means flying blind on feel — the thing native development exists to get right.

    4. Where will feedback hurt less? Your first version will be imperfect. Some founders want it in front of the most forgiving early adopters; others want it far from their loudest market until v2. Both are legitimate — but choose deliberately.

    Answer conflicts? Weight question 1 heaviest. Distribution is the hard part of mobile; build where you can actually reach people.

    The cost difference is real but small; the ceremony difference isn't

    The checkable costs: Apple's Developer Program is $99/year; Google Play is $25 once. Real difference, but if $74 changes your platform decision, the app is not the problem.

    The ceremony differs more than the price. Apple's review is famously stricter and the Mac requirement is absolute; Google's onboarding has grown its own requirements for new personal accounts (testing requirements before production release — check Play Console's current rules when you register, they have changed more than once). Neither is a reason to pick a platform; both are reasons to start the store paperwork a couple of weeks before you think you need it. The validation guide covers what to do while you wait.

    The 2026 reframe: it's a sequencing decision, not a bet

    Here is what actually changed, and why this question deserves less anxiety than it used to get.

    The old cost of choosing wrong was a rewrite: your iOS app was Swift, your Android future was a from-scratch Kotlin project (or a compromise — a cross-platform framework whose UI is native to neither). That made platform choice a bet with real downside, and the anxiety was rational.

    Shared-core architecture removed the downside without the compromise. Duet — ours; the mechanics are public — builds one project whose business logic is shared and enforced identical on both platforms, with real SwiftUI on iOS and real Kotlin + Jetpack Compose on Android as separate native shells. Start on one platform; when the second's turn comes, the core is already done and already tested — what remains is the native shell, not the app. Android on Modaal has been open to everyone since September 2026. {/ [GA-GATE] fallback: "Android on Modaal is in open beta — modaal.dev/android-open-beta." /}

    Practically, on Modaal the sequencing maps onto the plans: the free tier is one project with unlimited prompts on one platform of your choice — your question-1-through-4 answer — and Pro (€9/month billed annually) turns on the second platform and store distribution when your users demand it. Which they will tell you: the moment an app has traction, "when is it coming to iPhone/Android?" writes itself into your inbox, and that inbox is a better platform strategist than any article, this one included.

    The short version

    Your situationStart onWhy
    Subscription or paid app, US/Western Europe audienceiOSThe persistent monetisation rule of thumb — verify in your niche
    Reach-first, ads, global or emerging-market audienceAndroidRoughly two-thirds of the world’s devices
    B2B / internal toolWhatever the buyer’s fleet runsConsumer statistics are irrelevant here
    You carry the platform you’re building forThat oneDaily dogfooding beats any market report
    Truly undecided after the four questionsiOS first, Android fast-followMac requirement already met; second platform is a shell, not a rewrite — with a shared-core architecture

    Market share source: StatCounter, worldwide mobile OS, Aug 2026 — Android 67.61%, iOS 32.36%. Store fees: Apple $99/year, Google Play $25 one-time. No other statistics are asserted on this page.

    Frequently asked questions

    Follow four questions: where your first reachable users live (check per-country platform share — it varies enormously), how the app earns (paid/subscription leans iOS by the industry’s rule of thumb; reach and ads lean Android), which phone you personally carry (you must dogfood daily), and where imperfect v1 feedback hurts less. Weight the first question heaviest — distribution is the hard part.

    Worldwide, Android runs about two-thirds of mobile devices and iOS about one-third (StatCounter, Aug 2026: 67.61% vs 32.36%). But the split differs sharply by country, and revenue concentration differs from install concentration — so check the specific markets you will actually launch in rather than the global figure.

    Google Play costs $25 once; Apple’s Developer Program costs $99 per year — and iOS development requires a Mac. Real differences, but small next to everything else in a launch. The larger practical difference is process: Apple’s review is stricter, and Google’s new-account requirements have their own steps — start store paperwork early on either platform.

    The 2026 native answer is a shared-core architecture: Modaal’s Duet builds one project into two genuinely native apps — SwiftUI on iOS, Kotlin with Jetpack Compose on Android — with the shared logic enforced to behave identically. You can start on one platform (the free plan covers one of your choice, unlimited prompts) and add the second later as a shell, not a rewrite; Pro includes both platforms.

    Less than it used to. With platform-specific codebases, the wrong choice cost a rewrite; with a shared core, the cost of correcting course is building the second native shell over logic that already exists and is already tested. Pick using the four questions, ship, and let real user demand schedule the second platform.

    Start free. Ship native.

    One project, unlimited prompts. No card.

    Download for Mac

    Keep reading