mobile development, cloud computing and building great software

Rob Kerr

Addicted to coding since writing my first programs for the Commodore computers in the 1980s. Currently working as an independent contractor focused on native iOS development.




Should I build iOS or Android First?

Rob KerrRob Kerr

When considering delivering a mobile experience to our customers or end-users, this is probably the first question that comes to mind.

Perhaps you arrived on this page by Googling this very question! As of this writing, this question returns a list of over 22 million matches on Google!

Google result for this topic

That many matching pages probably gives a hint that the answer to this question is unsettled, to say the least. As with many unsettled questions, the answer depends on several factors.

The Bottom Line

I believe the right answer to this question is most dependent on who our app's target customers/users are, and what direct revenue contribution the app will makes to our product or service's success.

Do we know enough about our target users?

For one of our enterprise clients, we deployed a responsive mobile web application (i.e. device agnostic) for a "Bring Your Own Device" global user base. Then we measured precisely the proportion of iOS and Android devices via in-app analytics. Looking at what devices this audience was using, we saw the following:

Many of the this responsive web application's users still ask for a native version of the application. If we were to make this transition, the numbers make clear what the answer to the title question would be: we would satisfy the native requirements of 90% of end users by developing a universal iOS application first. The Android implementation could follow.

Does our application generate direct revenue?

If our application will generate revenue directly--for example if it requires a customer to pay before using it, or has in-app purchases--we need to consider which platform is likely to generate more revenue faster. This is especially true if we need the revenue from the first OS deployment to fund development for the second platform.

Consider App Annie's July 2015 retrospective, which reported that, while Android Play Store downloads are double the number of downloads as the iOS App Store, the App Store crates double the revenue compared to the Android Play Store.

App Annie Revenue by OS Stats

What makes this statistic all the more interesting is that it's often reported that Android is far and away the marketshare leader over iOS. And this is true--reports like this one from IDC tell us that Android holds 80% of worldwide marketshare.

IDC Market Share Chart

But how do we square the two statistics? Why does the iOS platform generate 50% of consumer downloads with only 20% of the market--and more importantly why does it generate double the revenue with only 20% of the market?

The logical conclusion is that iOS users are more likely to pay for software (apps) than Android users. Is this an important factor for our revenue model?

Where do our customers live?

Refer back to the worldwide smartphone market share chart. It tells us that 80% of worldwide shipments were for Android devices. Two things to really think about:

Refer to statistics published by ComScore in 2016 analyzing only installed base in the US:

ComScore Installed Base Chart

Remember, this represents US Only, and Installed Base. Here we see that Android still has more users we can reach, but by a much slimmer margin than the global IDC quarterly global shipment numbers might lead us to believe.

Should we always ship for iOS first, then?

All of this isn't to suggest that iOS should always be first. The main take-away from the market share and revenue share discussions is that we really have to think about where our revenue will come from. Are our customers US or Global?. Do we need App purchases/in-app purchases to succeed? Do we know our target demographic--and what devices that demo prefers?

While every situation is different, if you've wondered why it seems many DTC startups dependent on US customers spending money directly via their app start with iOS -- perhaps these stats provide some insight.

Our Team's Skills

Our team's skills may well be a factor in deciding which platform to ship first. Especially if we don't have the funding to hire additional staff or bring in outside help to build our first product version. If we have the skills on-hand, and especially if we're paying a fixed salary for those resources...then getting something to market to sell is better than not.

Why not use a technology to ship both platforms at once?

If money and time are no object, every client's answer to "Android or iOS" when I ask that question is: "Both".

There are essentially three ways that "Both" can be the answer to this article's question:

  1. We can fund the simultaneous development of native applications on both platforms
  2. We decide to develop a responsive web application instead of a native application
  3. We use a development tool that promises Write Once, Run Anywhere (WORA)

If our application needs a fully native experience to be successful and/or competitive, then option #1 is probably the best approach. If funding both platforms at once isn't possible, use the factors in this article to make a decision of which platform to deliver first.

Is our application idea really just a mobile version of our current web experience? Are there no offline use cases, no access to device-specific features (cameras, etc.), and is the app really for "content consumption" and not for "content creation"? Is budget a limiting factor? If so, maybe a mobile web site will meet the products needs.

Option #3 is the most difficult to evaluate. Using tools like Xamarin or PhoneGap-based platforms can produce user experiences identical to using native development tools and at lower cost -- if the following are all true:

  1. Our application requirements fit into what those tools do best and avoid requirements those tools don't meet as well
  2. The UI layer of our application is fairly simple, and should behave the same on both platforms
  3. There is a significant amount of non-UI code in the application. WORE toolsets return the highest efficiency for apps that have a lot of non-UI code


As I mentioned at the start, if the answer was clear-cut, there wouldn't be 20 million pages written on this subject! The answer really does depend on our situation. As outlined in this article, the factors that are most important in arriving at the best answer are:

  1. Who are our app's target users?
  2. Which countries do these users live in?
  3. Do we currently have any stats (i.e. from our web site) about what devices the target users actually use?
  4. Does the app need to generate direct revenue?
Rob Kerr

Rob Kerr

Addicted to coding since writing my first programs for the Commodore computers in the 1980s. Currently working as an independent contractor focused on native iOS development.