The App Developer Dilemma

A lot of great data came out yesterday that I believe is worthy of a deeper look. In my opinion, this data begins to shed light on some of the key questions I have had around platform engagement.

Flurry released some very insightful data that dug into the vast platform fragmentation across platforms. This data helps us draw clean conclusions around why it is very difficult for small to mid-size developers to survive if their goal is to have an app on as many devices as possible. To highlight this, Flurry makes the following point:

(for a developer) to ensure that your app is optimized to function well on 80% of the individual connected devices currently in use. How many different device models (e.g., Kindle Fire HD 8.9″ Wi-Fi, Galaxy S III) do you think you need to support? (the answer is) 156.

A developer would need to index and code for the different variables for 156 different devices just to cover 80% of the current connected devices in use. That sounds like a lot of work.

The Flurry data goes on to focus on more reasonable device coverage and estimated that if a developer simply wants their app to run on 50% of connected devices in use, it means supporting 18 devices. If you know anything about how app development, testing, troubleshooting, etc., works then you know this is a problem.

From the get-go my analysis has highlighted that developers would continue to commit the bulk of their resources to support iOS due to the minimal screen size and OS generation discrepancy that exists in Apple’s model. From the many startup briefings I am having with software companies in Silicon Valley, the iOS first mantra still rings true.

The next bit of data worth highlighting is around application engagement in iOS and Android. Flurry’s data highlights that even though Android has a greater number of people using the platform, iOS has a significantly greater application engagement level. iOS users engage an application 14 times more frequently than a user on Android.


Flurry even analyzed the data further and decided to look at application engagement not just of Android but by specific device brands running Android. The bit that stood out to me was the data around Samsung devices running Android. Flurry’s data returned that iOS (iPhone and iPad) users engage applications 7.7 times more than users on all of Samsung’s Android installed base.


This one has always been a puzzler for many of us because we constantly see the data (from many sources both public and private) telling us that iOS users are far more engaged with software on their smart devices. So with all the data pointing in this direction, we are faced with the question of why? I attempted to shed light on this with my column on iOS and sophisticated simplicity. My core conclusion is that iOS makes it easier to engage more with the software, but this point is subjective so let’s look at more data.

The Flurry data should shed some critical light on the development challenges facing many developers. The bottom line is for developers this is an issue of massive strategic proportions. comScore also shared some of their data, also targeted at developers, with the goal of highlighting some core differences between Android and iOS customers.

The main point I found interesting in the comScore data was their findings that the extremely high satisfaction rate of the iPhone leads to much higher device loyalty. Something I believe many in the media who write their famed “I’m switching” articles fail to realize is that the mass market simply doesn’t change for change sake. If they are happy and satisfied, then churn is rare.

Another bit of interesting data is related to the average income of consumers of both platforms.

Screen Shot 2013-03-06 at 7.10.38 PM

As you look at the chart above, you may be tempted to look at the lower part of the Android graph. That isn’t the one I want to focus on. I want to focus on the bottom part of the Apple graph. If you would have seen data going back a few years that looked at average income per platform, you would have seen that most Apple users skewed higher overall on the average income level. What we are now seeing is the iPhone growth, which led to 3.5% gains this last quarter, and caused Android to lose 1.3% market share, being a result of Apple growing their share of the lower end of the market. So what do you think will happen when a new iPhone comes out and the iPhone 5 becomes $99?

The Flurry article brings up an important point facing small to mid-size developers. Where should the focus their time and financial resources? This group is where true software innovation often comes from. Rather than spread themselves thin supporting a fragmented device universe, it seems wise that they focus on the customer base and platform which will reward them financially so they can keep innovating. This decision is of monumental strategic importance.

Published by

Ben Bajarin

Ben Bajarin is a Principal Analyst and the head of primary research at Creative Strategies, Inc - An industry analysis, market intelligence and research firm located in Silicon Valley. His primary focus is consumer technology and market trend research and he is responsible for studying over 30 countries. Full Bio

20 thoughts on “The App Developer Dilemma”

  1. Any sensible developer is going to do a quick Pareto analysis and go for the 20% of device types that deliver 80% of the likely market. This shows that the priorities of developers, and particularly their focus on iOS, are entirely rational.

    1. A small developer will do a Pareto analysis on the 80% , meaning 5% devices delivering 60% of market share. Strategic focus is even heightened.

    2. To Ben- Another insightful, smart article. Thank you.
      To Steve and Ben- Isn’t this the real reason “market share” for Android vs iOS needs to be specifically defined?
      For example, many years ago I wrote educational materials about certain illnesses for a company who sold these these to insurers for their plan members who had to live conditions that can be managed, such as asthma. Instead of pitching our wares just to the largest insurers we first targeted the insurers who had the largest percentages of plan members getting treated for asthma. The 6th largest insurer had the most plan members with asthma. It was an easy sell and a greater number of plan members benefitted (my kind of win-win).
      My point is “market share” is one thing but “targeted market share” may be where the attention should, no?

      1. To paraphrase Wee Willie Keeler, hit it where they are. We often take an excessive macro view of markets focusing on big-picture numbers when what really matters is what is going on in submarkets. For example, various iOS editions of the Merck manual cost $35 and up, expensive for an app. Their appeal to the general public is very low. But I suspect within the target market of health care professionals, the take rate is very high. The big problem is that it is very hard to get data on these important micro-markets.

      2. Yes and this is how its done in every consumer centric post mature industry. Think packaged goods, automotive, consumer electronics, etc. All marketing is heavily segmented. They know which types of customer they want and which ones they don’t.

        Apple has figured this out and others haven’t. Hence the throw it on the wall and see what sticks strategy we are seeing right now.

  2. any data comparing Samsung Galaxy SIII, Note 2 and iPhone 5 for active device index by any chance? I think that will really show the observers that iPhone really dominates the market.

  3. I understand there are over 4,000 distinct Android models in the wild. Market share can be market control or just chaos. I’m guessing Google wants to throw a big net and dump the rocks out later on. If Samsung doesn’t do that first.

    1. Yeah, but a lot of the distinct variety comes from China. These devices don’t have Google services on them so they don’t have the Play store anyway. So dev’s don’t actually have an avenue or really need to worry about those devices.

      This article that came out recently sheds more light on the China situation then even some of my articles have.

      1. Interesting. “The ChinaDroid supply chain also has big implications outside of China. All of the big foreign brands do their manufacturing in China…. China is truly the world’s factory for smartphones and tablets.”

  4. The difference between analyst and journalist shines brightly in this article. I keep looking for honest analyses such as this that show the demise of Apple is nigh, as spouted by a number a journalists; my quest continues.

    1. In defense of my sometime craft, I will note that most of the really bad journalism on Apple is lazy journalists quoting really bad analysts. Good journalists and good analysts produce good work. Bad journalists and bad analysts, who are legion, produce junk.

      1. Well said. I wish those who have proven worst would fade away but they seem to endure as weeds in a garden.

      2. Dang, I made that ‘all journalist’ mistake again. I just repeated an “all journalists are not bad” mantra and this time it will stick, Steve. It’s just that when I read good journalism, there is always evidence of good logic, sense and understanding in the author’s process, including good analysis, before putting pen to paper so I usually lump them in with good analysts.

        What I was attempting to say was that the work from this TechPinion site illuminates the topic and such becomes a learning experience. This can take place from both good journalist and good analyst.

        It is a shame, however, that the very respectable and important profession of journalism* has come under the control of corporate agenda. Where once we found many ‘Chris Hedges’ fearlessly writing for our newspapers and magazines, few opportunities remain today for good political journalists to express their findings; and this may be why journalism has lost some of its lustre.

        *without the power of good journalism, for example, can democracy survive?

  5. Comparing the huge disparity between Android vs IOS and Samsung vs IOS shows clearly, I think, that the vast majority of “Android” devices worldwide are in actual fact feature phones, not smartphones. Since Samsung makes both the Galaxy S3 and a bunch of cheaper models as well, I wish we had a third graph comparing “Flagship Android” to IOS as well — does the Android vs IOS disparity disappear if you restrict your dataset to Android phones that are actually as capable and versatile as Iphones, or is there still a gap?

  6. I don’t know if it is true or not, but I heard a comment over at Anandtech, that part of the “higher use” by Apple products is that Apple systems and Apps “call home” more frequently. I gather it had to do with switching between apps, and the system has to refresh and reload when you switch back and forth. So a lot of the higher traffic from Apple stuff is background admin stuff – in theory.
    Would be interesting to how much truth there was to that…

    1. That is interesting. I can do some digging and find out the degree that is true. But the data that theory can not explain is the much higher web traffic from iOS than Android.

  7. While I agree with Jim’s analysis of developers focusing their development on the most profitable market, I believe developers can use corporatecentral that puts their app on all major mobile devices (and web browsers). Mobile includes iOS, Android, Windows 8, Blackberry phones and tablets.

Leave a Reply

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