iOS first – a flawed strategy that startups have used for years – here’s why

by on | 21 Comments

Startups have been launching on iOS first for a long time, avoiding Android for as long as they can. Over the years, I’ve heard numerous explanations for that. And in the early days of Android, it made perfect sense: it had a smaller market share, it had less apps in the Play Store (then called Android Market), and it was the phone of ‘poor people’ – people that didn’t spend money on their phones. So let’s take a closer look at the reasons behind these decisions, such as developer revenue per app store, screen size market fragmentation, and OS fragmentation.

Android fragmentation – run in fear!

One “iOS first” argument has always been the much-feared Android market fragmentation. Android has hundreds of models of phones. They all have different screen sizes, resolutions, and pixel densities. Then, on top of it, the Android tablets also come in a million sizes and resolutions! Sheer madness!

Android versionsOn top of that, Google has been updating the Android OS constantly, and not every phone maker upgrades their existing phones to the latest Android version. This means that a lot of phones run older versions of Android, unlike iOS, where Apple sweeps everyone into the latest version.

Android fragmentation – screen differences are not so scary at all

In a recent article on Gizmodo, we are reassured that Android fragmentation is not difficult at all to code for, according to developer Russell Ivanovic’s detailed post. He calls it a complete myth, and tells us in lots of detail how Google tackled this issue very early in the life of the Android OS by providing some awesome developer tools that allow you to specify different UI layouts for different screen sizes (specifically tablet vs phone), and how pixel density differences are handled very elegantly. These tools have been part of Android from the beginning. Ask Android users if they have problems with badly laid out apps, and they’ll look at you confused – it’s rare.

This in is stark contrast with iOS – when the iPad first came out and lots of apps had ugly empty space around it because iOS had no way of coping with multiple screen sizes.

Android versions – painful to develop for?

Let’s make sure we’re not confusing two issues. When the phone in your pocket doesn’t run the latest version of Android, instant gadget envy rears its ugly head, of course. And indeed, there are lots of versions of Android out there. But what does it mean from the developers point of view?

Coding for Android – a simple choice

Android is divided roughly into two main versions: Android 2 and 4. If you want to build an app that works on version 2 AND 4, you have to take the lowest common denominator and code for version 2, and it will run fine on both. If you want an app that only runs on version 4, allowing you to use the latest and greatest coding techniques, the version 2 people cannot run your app anymore.

How is this handled in practice? 14.2% of Android users are still on version 2. These are mostly people in emerging markets. So in practice, everyone uses the latest coding techniques and codes for Android 4 only, leaving behind some users in countries that would never pay for anything anyway, plus version 2 is losing market share rapidly.

Is this really true? I’ve talked to lots of Android and iOS teams, and they say if they implement the same app on both platforms, the development time is roughly the same, and testing on Android is mostly a non-issue, as you can do it on your computer in emulators.

Metrics revisited

We all know that Android commands a bigger market share in the global mobile (phones + tablets) market, right? But not where it matters, in the US? Wrong – Android has 61.9% market share in the US, whereas Apple has 32.5%. So for every iPhone in the US there are two Android phones in the US, roughly. Go to China, and Android’s domination is even stronger at 82.7%, and in Europe, 73.3%. Those are some pretty big numbers.

So – let’s get this straight, even in its strongest market, the US, Apple is outnumbered 2:1.

The final metric: MONEY spent in the app stores

Since this is the only metric still pointing in favor of an ‘iOS first’ strategy, this metric is constantly all over the media, most recently by Andreessen Horowitz analyst Benedict Evans. Part of the problem is that both Google and Apple release very few numbers that give us a detailed picture. At the recent Google I/O conference Google said it paid $5bn to developers over the past year.

Google and Apple developer payments

Google and Apple developer payments


Jan Dawson, a Google I/O attendee, did his magic, and concludes “Google is catching up quickly in payments to developers”. If you look at this graph, in less than 12 months, Android will dominate in this final metric.

But did that the total revenue number even matter?

We’re so focused on revenue from the app store, that we ignore some facts: lots of apps support other platforms, and generate revenue on other ways. My phone has a great SalesForce CRM app on it – it’s free. But of course our company pays for the CRM! Apps also often drive revenue on other platforms. In Android apps, you can freely send users to another platform to do a transaction – how much revenue is generated that way? iOS blocks this, by the way, and insists on a 30% cut.

Other apps are ad based, and that revenue can come from any one of hundreds of ad exchanges. Again, none of that revenue is part of this graph.

So why is there all this irrational decision making by startups?

Apple is beloved in the startup world and the tech press for good reasons: we all love our Apple gear! I can’t remember the last time i met a tech journalist or a startup founder toting anything other than a Macbook, an iphone, and maybe an iPad mini. In San Francisco, iOS marketshare must be 80% or higher, if you just sit on a MUNI train and look around. But wait, i thought the US marketshare number was much lower? That’s right, it’s 32.5%. So why are all my friends and the people i work with iPhone users?

I live in a tech bubble, and so do you probably

The truth is, if you work in the tech world, you’re not around average people. That’s probably true for any industry, but in this case, the startup founders who have to make the decision with what mobile technology to go first, are people who’s parents probably have iPhone’s too, let alone their friends, their co-founders, their investors, their developers, and their kids. This creates a very biased view of the world. Even when we do customer validation (remember the Lean Methodology!), we ask the people around us, or in the street of the city where we live. In the San Francisco Bay Area, this inevitably leads to a very skewed view.

Conclusion: let’s think rationally and make objective decisions

When choosing to launch on iOS or Android, you’re probably under a lot of pressure. Take some time, leave San Francisco for a day or two if you live there, and do some independent research. This article contains some links to hard data, use them as a tool to explore! But don’t let your investor or your developer pressure you – business decisions need to be made based on fact, not emotion.



Written by Wessel Kooyman

The author, Wessel Kooyman, is the CEO of Cole Street, a software company. Wessel is an entrepreneur and developer and has a lot of expertise in working for startups in Europe and the US.

21 responses to “iOS first – a flawed strategy that startups have used for years – here’s why”

  1. kosmo says:

    All I know is that my $3.99 iOS app outsells my $3.99 Android app 5-10:1.

    • Ben says:

      Yup same here. iOS wins every time.

      • ashafa says:

        This is usually the product of trying to mimic your iOS app on Android. Porting over the look and feel of iOS apps on android without taking into account the plethora of advantages the android OS has over iOS leads to inferior looking and performing Android applications.

        • andyhmltn says:

          Nonsense. That wouldn’t fix a clear 5-10:1 sales ratio

          • Oletros says:

            Clear? Have you seen the number of downloads?

          • users aren’t stupid.. they want apps that look good on their devices, not apps that look out of place or perform like junk because they are just ports of an iOS app.

          • Adrien says:

            It’s quite simple : users don’t pay an application on Android, it’s 90% free, so for 4$ your app nedd to be a veeeeery good one to be bought. On iOS, 4$ is almost free cause the average price for an application is higher.

            So, use some adds in your abndroid app, make it free, and it’ll a 10:1 ratio for android.

      • Oletros says:

        Can you link to your applications on Google Play?

    • Oletros says:

      Can you link to your application on Google Play?

    • So lets see a link to the app so we can see if it uses Android Design guidelines or if it’s just a crappy iOS port.

  2. Paul Sawers says:

    Interesting stuff. However, a few issues.

    If we’re talking about the trend over the past few years, then yeah startups have definitely been more inclined towards iOS first, but over the past few years iOS has typically had a much larger market share in key markets. This ‘32%’ you speak of in the US isn’t market share, it’s the number of devices that shipped over a certain timeframe, in this case a quarter. It may or may not represent a future trend, but it doesn’t mean that only 32% of Americans have an iOS device in their hands. In terms of US market split, iOS typically has has had more than 40%.

    When startups decide to build mobile software, it makes sense they focus on one platform if they have limited resources. Now sensibly, that would be iOS or Android. As you note, most devs, especially in the Valley and similar places, are likely iOS users themselves, so it does kinda make sense that they have hitherto focused on iOS – something they understand better and it’s a big market.

    Also, many studies have shown that iOS device owners are more inclined to pay for things – be it in-app purchases or otherwise. It’s been easier to monetize on iOS. Demographics are key here.

    Also, fragmentation isn’t as big a problem for devs as many folk make out, but it’s still a problem. Even with some of the big-name apps to this day, you’ll have an app’s feature that works flawlessly on one phone, but not so on another, even phones with the same version of Android installed. These are usually ironed out over time, but the initial teething problems are still an issue. Bluetooth Smart, for example, is still pretty limited on many Android devices.

    When you throw all these things together, I wouldn’t say that it’s a ‘flawed’ strategy that startup devs have typically focused on iOS first before looking to expand to other platforms. They have limited resources, therefore look to the simplest platform to build for, one that they themselves understand, and one that has a very big market share. It’s not purely about “which platform has the most number of users? Android? Cool, let’s go with that”. There are a myriad of factors that come in to play.

    It’s worth noting here that this is speaking purely from a historical perspective. Android has evolved hugely in recent times, and is continuing to, and we will likely see startups put as much focus on, if not more so, on Android than on iOS.

    • Adrien says:

      Last numbers indicate that iOS will decrease to 15% of the smartphone market before 2016, and 18% of the tablet market before 2017…

      American market is a small one if you compare it to the global one.

  3. Ron says:

    You quickly dismiss fragmentation — but you don’t consider at all the increased costs of the QA matrix for AOS devices. It gets expensive very quickly. And when you try and keep those costs down by limiting the number of devices supported, there is typically a huge outcry from the community. You also don’t consider the subtle differences in the lowerlevel stacks, like BLE, between devices and AOS versions — we’ve been killed by that issue where I work. Yes AOS is typically worth the hassle, and is becoming more and more worth it, but you have to face the truth that many aspects of development for AOS are much more expensive than for iOS, and it’s rational to pay attention to those expenses.

  4. JSingh says:

    Try developing a great app for Android on that terrible eclipse and even more terrible that emulator. Then make that same app in ios using Xcode. You will delete your article then.

    See developers are not idiots who follow herd mentality. They are very smart. They choose the least effort with great output. That’s why ios.

    And fragmentation doesn’t matter if you make a crappy app. So also include quality of apps produced in your article.

    • Ludjer says:

      IF you dont like eclipse the use gradle to do your builds. Also Android Studio or intelliJ is the best IDE by far for android development(even google is moving the default ide from eclipse to android studio). Even though the most popular is eclipse lots of people do not use intellij because they think it costs money, which is not true. It has a community which is free and lower amount of features and has the ultimate one which is feature rich. Though google is beta testing a android studio which is based on the community edition of intellij idea.

      With regards to the emulator you should chose the X86 emulator and then everything will be smooth. Though if you are a serious app developer you should never use a emulator. I personally use ADB over tcp(wifi) and test directly on the phone. Its one button to build deploy and launch on your phone and gives you much better feed back since you see how it would look on a phone and still have short build cycles.

  5. Melinda Green says:

    For apps intended to make money directly, I’ve been promoting an Android-first strategy for a long time. This may seem counterintuitive because Android users largely expect apps to be free. My reasoning is that you can essentially use Android users as beta testers. This works especially well because your updates are available in the Play store in a couple of hours, whereas iOS updates typically require 2 weeks. So you make all of your mistakes on Android, and then once you’ve got an app that you know users find valuable, then and only then do you port to iOS in order to cash in.

    I’ve only had partial success promoting this strategy, and because of this article, I think I now know why: The decision makers who control the roadmap are Apple users and want to see it on their own devices as soon as possible! From their perspective that makes sense, because that lets them weigh in on the UI where everyone feels they are expert.

Leave a Reply

Your email address will not be published. Required fields are marked *