Fantastic article about design on Android by Cennydd Bowles, design lead at Twitter.
Android design is indeed more difficult than iOS design in that it offers fewer constraints. But any skilled designer can handle that with a bit of effort. My uncharitable interpretation for this class of responses is simple laziness, and if Android forces designers to drop a pixel-perfect mentality and adopt approaches that suit a diverse world, then that’s no bad thing.
The evidence is out there for all to see. Android developers – developers who are Android-focused instead of iOS-focused – come up with absolutely beautiful Android applications all the time. I have no doubt that it’s harder to do so on Android than it is on iOS, but the cold and harsh truth is that there are also a hell of a lot more Android users and devices out there. If your iOS application requires two full-time developers, is it really fair to expect your Android application to require the same, even though the user base is four to five times as large?
A translation consisting of 3000 words takes me about a work day. A translation of 12000 words takes me four work days. None of my clients expects me to translate 12000 words in the same amount of time as 3000 words without a serious degradation in quality.
Bowles also dives into the argument that Android users are less willing to pay than iOS users.
Socially, excluding Android users seems almost prejudicial. Unlike Android is difficult, this isn’t about about mere convenience; it’s a value judgment on who is worth designing for. Put uncharitably, the root issue is “Android users are poor”.
If you are an iOS developer, and you port your Android application over as a side-project, is it really so surprising that Android users aren’t buying your application? Could it simply be that your potentially poor iOS-to-Android port simply isn’t even worth paying for? If you do not develop and design with Android’s strengths in mind, Android users won’t be as willing to pay as your iOS users, the platform whose strengths you do develop and design for.
I translate English into Dutch, and since this is my speciality, I’m pretty good at it and my clients are willing to pay good money for my services. I could also translate German into Dutch, but since my German isn’t nearly as good as my English, my clients aren’t going to pay for it. I can translate German into Dutch just fine, but the quality will be far less than my English-to-Dutch translations.
Even then, Android’s userbase is far larger than iOS’, so even if only 50% of your Android users pay, and 100% of your iOS users (unlikely figures), Android still provides a more worthwhile revenue stream.
Still, the core issue is that Android is a different platform and ecosystem than iOS, with different strengths and weaknesses, and as such, requires different talents and mindsets. Translating English is different than translating German. I realise that. Developers should realise the same, and understand that being a good iOS developer does not make you a good Android developer – or vice versa.
Uh…yes, that is a perfectly rational expectation. The fact that Android has more users does not mean you need more developers in order to create an application. It might be needed for other reasons but number of users is not a factor.
You’re missing the part about more devices.
“More and different devices” is not the same as “More users”. If you mean because of the more varied devices then say so.
…but I do.
If this:
“If your iOS application requires two full-time developers, is it really fair to expect your Android application to require the same, even though the user base is four to five times as large?”
Read as this:
“If your iOS application requires two full-time developers, is it really fair to expect your Android application to require the same, even though there are hundreds of more devices with no common hardware standard?”
It would be a lot easier to accept your point… but then, it would just be what iOS developers have been saying for 5 years now.
It would also be nice if the author (or yourself) displayed the same graph with Google Android separated out from AOSP (which Horace posted shortly after). But then you’d have to account for the actual addressable market being much smaller (or more difficult if you are actually suggesting developing localized Chinese apps and promoting them through a multitude of Stores rather than just Google Play).
If your sole point is: specialized development is better than generalized development and to do specialized development for more than one market requires more resources and maybe more than double the resources, there’s not much controversy, or even a point, there. However, the question remains: why should someone develop for Android rather than iOS (or devote more resources to Android rather than iOS)? You (nor the original author) didn’t even remotely start to broach that topic, never mind answer the question.
In other words, this sentence appears to be your true thesis:
“Even then, Android’s userbase is far larger than iOS’, so even if only 50% of your Android users pay, and 100% of your iOS users (unlikely figures), Android still provides a more worthwhile revenue stream.”
But that statement is laughable and you did absolutely nothing to support it.
Edited 2014-03-24 16:48 UTC
Look at Windows phones and iOS phones. Do you really think that there is really only one platform to support? Have you seen that the number of devices is increasing in both non-Android camps too? What situation will be there in next two or three years? Something like 10 cheap half-broken-down iPhones, something like 10 different Windows sizes (because Android has them too)? Did you notice, that Microsoft and Apple just dump old phones and old versions of systems with them? What about your customers? There will be a huge market for special apps for bussines in next few years. You will not be able to stop support app for you client only because Microsoft or Apple will tell you that some of devices your customer bought are obsolete now… You will end up in the same situation with Apple or Microsoft you see in Android now. Every platform has disadvantages. More open platform has disadvatages too, but if you know enough, you are able to overpass (see Android for HTC HD2, CyanogenMod for original Samsung Galaxy S etc).
Not in that sentence. If you had used a term like platform targets, or whatever Android’s nomenclature is, instead of “user base”, it would have been much clearer.
English is hard. What you’ve written, “more Android users and devices”, reads like “more users” because it doesn’t mention any variety in the devices. It seems repetitive because of course more users would need more devices. The next sentence isn’t helping because you only mention the user base. I assume what you meant to say is something like “You need more developers or more time for Android because there is a larger variety of devices to support.”
I think a lot of it has to do with the app economy being majorly fragile in the first place. Uncertainty beyond a reasonable amount can keep a company from taking the dive and making the port.
These are sometimes dollar apps which people are reluctant enough from purchasing on iOS, let alone Android. Its difficult to justify both a demographic with smaller purchasing power and a platform which requires more engineering resources by virtue of its diversity.
Android’s reach works against it in some cases, for example, in many countries where there is no carrier billing with Play.
Then there’s the ugly piracy issue that rears its head, I’ve seen enough developers moan about their apps being pirated to know that this is a pretty common issue.
Google can do a lot more about fixing this problem than it’s currently doing. It could expand carrier billing, protect company IP, improve their tooling / store front, improve discoverability, make revenue guarantees. Etc.
There’s a laundry list of reasons why it doesn’t make sense to develop for Android, and the majority aren’t dev related.
Edited 2014-03-24 12:16 UTC
Out of interest, do you have any evidence of this apart from “developers moaning”?
In one particular situation, I saw a developer complain about the disparity between the numbers of logins or check-ins for their app compared to the number of sales. However it turned out that the disparity was 100% explained by the fact that the users had multiple Android devices. The developer just simply hadn’t taken that into account and assumed the massive disparity was due to piracy.
A google search returns innumerable results with varying degrees of writing quality, evidence, arguments for and against, etc but it seems a somewhat evident fact based on a number of factors:
1. Android market share leads in markets with historic trends of greater software and content piracy (of all kinds) based on numerous economic and cultural factors.
2. Observable attributes akin to “the law of large numbers”: as market share increases, desirable attributes tend to decline while undesirable attributes tend to increase: these attributes can be education, wealth, willingness to pay, rate of upgrade, etc.
3. Apple’s interest to minimize “piracy” versus Androids interest and ability to minimize it.
*Piracy on iOS is essentially: jailbreak + Cydia + Installous alternative. That’s it. Apple’s efforts to stem piracy go beyond control, curation, and removal from the App Store; it’s also born out in Apple’s basic security needs — efforts to prevent hacks that enable jailbreaking directly prevent the potential for piracy. (Technologically-capable users certainly wouldn’t say that piracy is much more difficult on iOS, but for the average user, the hassle of staying on older OS versions, keep track of working jailbreaks, the bugs of jailbreaks, which hardware and software combinations are supported, whether or not jailbreak-only software or “pirated” software is worthwhile, etc. minimizes the extent that a larger % of iOS users are pursuing piracy.) Apple has, very generally speaking, one vector to piracy and is very much motivated and capable of minimizing that vector.
*Google may have some similar motivation and means to prevent piracy on Google Play but they also have competing interests and a complete lack (and lack of desire to) prevent several vectors (ROMs, alternate stores, competing OEM ecosystems, side loading, etc) that more easily enable (and sometimes even promote — with a nudge, nudge) piracy.
*Even where capabilities and motivations align between Google and Apple to stem piracy in their app stores, it seems Apple actually does more and is more proactive — even though both could do a great deal of improvement. (As one small measure of this: I think it’s safe to say you are more likely to find literally the same app, provided by a different developer — exact same code with minor changes, likely still drafting off of design, color scheme, and naming trademarks of the pirated app, with different pricing or alternate ad sources — in the Play Store than in the App Store.)
I could go on, but those 3 factors make it reasonable to me that piracy is likely a greater problem for Android than iOS.
However… I couldn’t find the one story that you cite. Could you provide a link to the initial post and the follow-up correction? Any articles that you think contain strong argument and strong evidence to support your own perspective or either side of the argument? Thanks.
Although I think it’s reasonable to say that piracy is greater on Android than iOS, I would still say: I don’t think we have good metrics never mind good, like-for-like comparisons between competing, quality metrics. Nor do I think piracy alone is necessarily a substantial enough factor to dissuade developers. Nor do I even know how it compares to other factors and impediments that developers must face — pros and cons all. But… I would need a whole lot of evidence to dissuade me from thinking piracy is greater or more problematic on Android than iOS. And so far I’ve found any evidence of the kind completely lacking.
Edited 2014-03-24 22:17 UTC
I wasn’t attempting to say that Android has less of a problem with piracy than iOS. In fact, I didn’t even mention iOS. I was specifically asking if there was anything more than just hearsay about the Android piracy “problem”.
No, it doesn’t. This is exactly what I was _not_ asking for. Providing some sensible (and some less sensible) reasons as to why there _might_ be a piracy problem, does not mean there _is_ a piracy problem.
It was a conversation on Reddit. It wasn’t meant to be empirical evidence of anything. Rather I was presenting it as anecdotal evidence that I hadn’t actually come across anyone that had an issue with piracy on Android. Only someone who thought they had, but actually didn’t. Everything else I had heard had been from second hand sources (such as yourself) suggesting there is a problem, but not backing it up with any hard data.
I was lazy and didn’t do a Google search as you suggest. I now have, and have read a couple of articles. On the face of it, it does sound like a problem. For example, http://www.wired.co.uk/news/archive/2012-05/02/android-market-game-… suggests that even accounting for every user having two Android devices, they still experienced a piracy rate of 83%. As an anecdotal counterpoint to the article, I have actually bought Football Manager Handheld (the app in the article), and have already installed it on 4 separate devices, additionally I’ve installed more than once on some of the devices (due to me changing/updating ROMs). So rather than 2 installs, for me at least, they should have used probably 6 installs. I understand this doesn’t prove anything; I’m just trying to show how difficult it is to actually work these things out properly, and that jumping to conclusions sometimes creates a misleading picture.
Other articles, such as http://www.androidpolice.com/2012/07/31/editorial-just-how-bad-is-a… at least mention that some consumers aren’t able to buy Android apps (whether in-app purchases or not). The main example is people in China. I understand that there is sometimes a marginal cost to piracy (such as the bandwidth required for the downloading of extra content as described in the first article), but giving a percentage piracy could be considered misleading, since it suggests people are pirating it _instead_ of buying it, even if it simply isn’t possible for them to buy it (I understand that is still literally piracy, it’s just that it doesn’t affect the revenue of the company, although potentially the costs).
The reason I was asking for more evidence is that, for example, Google has systems in place to help prevent piracy, such as https://developer.android.com/distribute/googleplay/about/distributi… and https://developer.android.com/google/play/billing/billing_best_pract… . Do the companies that claim problems with privacy use these tools? Is there a problem with them? Do they not help privacy at all? Do they help only a little?
I just haven’t seen anything that answers these questions.
100% agree.
As an aside, you make some amazing claims, and on some things I just don’t understand your thought process. Would you care to explain:
I don’t actually get what you’re saying here. It sounds like you’re saying that because Android has large usage numbers, it has a negative effect on education, wealth etc?!
and
Really? What possible reason does Google have for enabling piracy? Contrary to what you say, Google does not support custom ROMs: as well as voiding warranty to install them, ROMs can’t include Google’s apps due to licensing ( http://wiki.cyanogenmod.org/w/Google_Apps ) and I believe that installing them yourself it at best legally dubious. Google also does not support alternative stores and AFAIK doesn’t allow them to be submitted to the Play Store. Google also hates the current extent of the OEM ecosystems, hence why they made a deal with Samsung recently to “bring the Samsung version of Android more inline with Google’s vision” ( http://www.engadget.com/2014/01/29/google-and-samsung/ ). In the latest versions of Android, Google have also hidden the option to be able to install non-Play Store apps (i.e. side-load). Un-hiding is relatively easy (and is necessary for developers), but it’s a small sign that Google doesn’t like side-loading.
I’m not sure if you are strictly talking about individual piracy here, or whether you are also ranging into people committing copyright and trademark infringement (someone has made a copycat app), but in either case, Google does not allow (and I have never seen in my years of use) either type of app in the Play Store. I have no idea how you think piracy occurs on Android, and to be honest, I don’t know myself since I’ve never seen it, but I am very confident in stating that it does not occur through the Play Store.
Technologically, side-loading on Android is not too difficult if someone gives you instructions, but if the developers use the Google Play Licensing (as I linked to above) that would prevent side-loading unless a hacker actually modifies and cracks the app. That’s not necessarily easy to start with, but other techniques such as encryption (also on the developer page) and obfuscation help prevent that. In other words, if developers use the libraries provided by Google (for free), piracy really shouldn’t be a big problem. So what is going wrong?
stimut,
This link seems to paint a picture where the worldwide copyright infringement stats are _very_ highly skewed.
http://www.androidpolice.com/2012/07/31/editorial-just-how-bad-is-a…
It’s anecdotal evidence from a miniscule data set, nothing quantifying as “proof”. However assuming the pattern holds then it seems to be a completely plausible and logical explanation for the high infringement rates we’re seeing. In places where paid apps are not available in google’s play store, well the infringement rate will necessarily approach 100% and in such circumstances I wouldn’t blame them.
Does anyone have access to infringement data broken down by country rather than just total figures? It would be very interesting to compare the infringement statistics with and without the countries where paid apps are not offered.
Edited 2014-03-26 04:59 UTC
Yes, I agree. You’ve basically summarized that part of my argument.
All of this is why I asked Nelson if he had any hard evidence, or if it was just “developers moaning”.
Judging by the fact that no one can point to anything like this, I suspect that although piracy is happening, it looks like it is in places like China, where consumers aren’t able to purchase the apps, so developers aren’t actually losing sales.
Of course, if Google was able to sort out whatever the problem is that prevents them from selling in China, then that would be a different story.
I don’t care what you were trying to say. I am stating: Android has a greater piracy problem than iOS. As long as iOS remains more profitable on a relative basis than Android, to me, it is sufficient to say that Android has a greater piracy problem than iOS. If it’s YOUR CLAIM that piracy exists but it’s not a problem, it’s YOUR OBLIGATION to PROVE so, not just to say you don’t have any evidence (nor have you looked for it) to support one argument or the other.
Which is not what I am saying: I am saying that it’s almost unquestionable that piracy is more problematic on Android than iOS. If you can prove otherwise, or even simply present some contrary evidence, I’d be happy to look at it.
This is my point regarding your comment: you used your complete and utter laziness as evidence. Not looking for evidence is not a basis to claim there is no evidence.
Of course it will affect sales, even if people literally couldn’t purchase it. Google could provide payment options: are you suggesting that no one would pay with a payment option? Do you think people willing to pay are more or less willing to pay after getting things for free for more than 5 years?
Seriously? Are you asking if these “tools” prevent piracy or not? Hahaha! Everyone has tools to limit piracy. My whole point is that Google has far fewer tools to prevent piracy.
Gladly, since there’s nothing particularly “amazing” about them.
More or less. With greater market share, the demographics become more average, less desirable. Yes.
Google doesn’t have to “support” any of these means; simply having a policy that permits these vectors is sufficient to qualify as more lax (or “encouraging”) than Apple. Does Google close hacks to allow loading of ROMs? Do they try to prevent the distribution of ROMs? Then they are more encouraging of piracy and have more open vectors to piracy than Apple does.
I’m talking about piracy by repackaging. That is, piracy by other developers and distributors rather than the end user. The effect is the same if not greater. I can assure you that it has occurred on the Play Store and that there are numerous alternative stores that are largely based on the practice.
It is occurring quite easily and quite frequently. This is fairly obvious. What is “going wrong” is your thinking apparently. If you think piracy is difficult, limited, or that developers shouldn’t give it much thought, you don’t know the landscape whatsoever.
Edited 2014-03-28 16:53 UTC
There is no carrier billing for the vast majority of users, no matter what they use.
No paid apps in many countries and credit card requirement – is the bigger issue for Play.
And I’ve seen even more that deal with those issues quite well.
Carrier billing is available only in a few countries even for AppStore, so please…
IP protection is provided as much as they can, they can’t control anything outside of their properties.
Discoverability is as big an issue on AppStore as Play, there is literally no difference by now – both are crap.
What is revenue guarantees? Pay money for apps that no one is buying?
Edited 2014-03-27 10:50 UTC
You could say that, but there is pervasive carrier billing for other platforms. Windows Phone has carrier billing in 34 countries.
Its amazing that you’ve seen more instances as compared to a figure I never divulged. Piracy IS an issue despite the collective head in the sand attitudes of some here.
iOS has a different demographic, they’re strongest in developed countries, major markets.
And that’s understandable, but piracy on Android by virtue of being Android is still a Google problem. It doesn’t disappear because you come here and say they’re trying hard.
I wouldn’t say no difference, but honestly when your demographic is a harder sell in the first place, you need to try harder than Apple.
Yes, or protect porting investments. Guarantee revenue if its over a threshold, but less than a cap.
BBRY did it nicely by guaranteeing 10k if you made 1k in revenue. Nokia pays porting and marketing expenses.
Microsoft will damn near port the app for you by throwing its developer resources your way.
And as Apple has shown it’s largely irrelevant.
It’s amazing that you don’t see the irony in my statement.
Piracy is a problem that companies deal with. It’s not the reason that some companies doge Android development. It’s not even one of the top reasons.
That is the main reason for iOS’s ecosystem success. Where piracy and other things are largely irrelevant.
Harder sell and discoverability are not the same thing, thus you see that the majority of revenues in app markets are taken by the top X developers.
BB and WP are not in the same league as Android and iOS. They need to get developers to their platforms, so they will do anything. No matter how many developers doge Android, Android is not in the same boat as BB or WP.
Come on dude, your text show no evidence what so ever. Every research shows the otherwise, that iOS users pay more for apps.
It may change in the future, but right now developing for iOS gives you more return ($$), and less pain (debuging, testing, etc).
Give me real data showing otherwise and maybe I’ll chance my mind. But right now all your “evidences” are shenanigans.
IMO a great percentage of Android phones are bought as feature phones that also have the ability to play Angry Birds, Asphalt 8 and Temple Run and run free apps (like apps from TV stations, social services and gas station finder apps).
I am not saying said phones are junk. To the contrary, they usually have quad core CPUs, 1GB ram and other goodies. What I am saying is that the people buying them aren’t invested enough on mobile games or mobile apps to do a purchase.
I know, because my dad, mom and sister have an Android phone. They use a lot the camera, radio and music and they use free games and apps, but haven’t purchased a single game or app. They don’t care about mobile gaming enough to buy games, and their mobile app needs are covered by free apps. Only I buy stuff. That’s 1 out of 4.
Instead, when you shell out 600 or 700 euros for an iPhone, chances are you are invested on mobile gaming and mobile apps enough to do a purchase. So almost every iOS user is a payer.
So, if you port to Android and expect 4x as much sales (because Android has 4x the market share of iOS), you are going to be dissapointed.
PS: Another problem is Google Play still requires a credit card in most countries (aka there is no Google Play Store reload card), which is something not everyone has outside the US. I had to go to a bank and issue a prepaid card to buy from Google Play.
Edited 2014-03-24 13:23 UTC
Do your banks not issue Debit cards with a VISA/Mastercard attached? In the UK, every Debit card is backed by either VISA, Mastercard or similar. The money still comes directly from your bank account, not a “credit line” and you can’t spend more than is in your account (including overdraft), but anywhere that accepts VISA etc, will accept your card. I can’t believe/don’t understand why this isn’t common place elsewhere.
It’s been a while since I set up my payment info with Google Wallet, but IIRC they require a credit card. A long time ago I tried linking my debit card to PayPal and that didn’t work either. Guess it depends on your bank’s usage policies.
I own a small software company that develops apps for enterprises. When I started in 2011, all client requested an iOS app, some of them a port to Android. In 2013 basically all my clients requests apps for both platform and most of them order them simultaneously. In January 2014 I received a first customer request for an Android only app. You can say that this is an anecdote, but I do see a shifting trend.
Naturally an iOS development specialist in most cases performs iOS specific development tasks more efficiently than a Android development specialist or an generalist. That is why he/she is called iOS development specialist.
Other way round an Android development specialist in most cases performs Android specific development tasks more efficiently than an iOS development specialist or a generalist. That is why he/she is called Android development specialist.
A development generalist is able to do an iOS as well as an Android development task. The generalist is in most cases not as efficient as a specialist, but the resulting code is — in my 15+ years’ experience — in most cases better structured and more portable than specialists code.
So, is a development generalist is able to do an iOS as well as an Android development task, well structured and very portable both a good iOS developer and a good Android developer?
Greetings,
pica
The number of units sold or number of device users is utterly irrelevant to commercial App designers. All that maters is how much revenue and profits can be generated by a given amount of effort. At the moment, mostly, developers can be confident of earning a significant amount more from a fixed amount of cost/investment from developing for iOS than for Android. Similarly if you are developing an app to add value to a non-app business then the easiest target to hit to get the most active users for the least effort or resources is iOS.
None of this is an accident but a reflection of the different priorities, designs and strategies of Apple and Google.
None of this matters much, the ecosystems of both iOS and Android are healthy and will remain so. There are no disadvantages for anybody associated or invested in iOS from being a minority platform just as there are no advantages for anybody associated or invested in Android from being a majority platform. Mobile devices are not like the old PC market.
You’d have to be hopelessly shortsighted to believe that the units sold/in use has no bearing on the revenue/profits that can be earned by developing for that platform. The value of a platform to a developer could be expressed as the amount of revenue they’re likely to make from each individual user multiplied by the userbase. If Android continues growing (and outpacing iOS) at the same rate is has been, then there will inevitably be a point where that equation no longer favours iOS – just by sheer numbers, even if individual iOS users continue to spend more on apps.
And at that point, why would app developers continue to favor iOS? (Hint: they wouldn’t)
The logic that you need more developers is flawed. You don’t necessarily need more developers for an app, but you might need more time to do proper testing.
Most of this could be mitigated if Google would grow a pair and make some decent hardware guidelines for android branding and google play access.
I would also make an argument that it’s easier to code for android because of the programming language difference with iOS.
Huh? I had a look at Google’s compatibility definition, looks pretty detailed, precisely to make sure OEMs can’t mangle important APIs, so an decently-made app decently can run on all phones having the version it was coded for.
The problems with Android development are:
1)Version fragmentation. Want to use that cool new API the latest version of Android introduced? Bummer, you have to bend backwards and backport to older versions if you want to get more that 10% of users. Compare and constrast with iOS, where 80+% are running the latest version.
2)The NDK is barebones. Really barebones. If you want to put hardware-accelerated video or 3D in your app, you are going to hit a wall with the SDK. So, you have to relearn a completely new thing (the NDK), which is hard to learn and diffcult to use too. Compare and contrast with Apple’s single SDK and slick Xcode. It’s not a coincidence most developers complaining about the Android developer experience are game developers.
Other things like handling screen sizes and different SoCs are non-problems if you know what you are doing. But there is no way to dance around the fact version fragmentation and the barebones unfriendly NDK are problems.
Edited 2014-03-24 23:17 UTC
The NDK is a joke, compared with the iOS, Windows Phone and Blackberry experience.
It is bare bones, Google looks at it as a stepchild and even APIs that are intended for game developers like the Game Services are only Java.
Only now, after too much pressure, did they release a NDK version of it.
Even though they make use of GCC/Clang, they cripple them to C and C++. There are a few open issues to offer the full set of languages supported by those compilers.
Then there is the developer experience. It is hard to believe that a company like Google is not able to offer, since Android 1.5, a working debugger.
Again the move to Android Studio, left out the NDK developers making use of Eclipse CDT, with no upgrade path.
Not that Java support is any better, with Google’s fork stuck on Java 6 level (Java 7 only with 4.4) and Dalvik GC/JIT leaves a lot to be desired. Everyone is looking what ART might bring to the picture, besides AOT compilation.
it is far cheaper to create a fixed screen design, and to wow with such design.
On the other hand few designer are making flexible width/ resposive website, because it is far more difficult, because graphic tool are not designed for it.
the fixed frame philosophy makes design easier. And also make feedback resolving easier (you can move the button 1 pixel to the left satisfy the producer).
Returning to the Android, the variety of aspect ratio makes it difficult to make a “pixel perfect” design (as described). And this also concerns the user experience.
Setting Aside design issue, I would also point that unfortunately Android user are too diversified to satisfy, and it is far easier to draw critics from android user than constructive comment, and unfortunately I would add that the robustness of the appication (due to various factor independent from the developer) does not help. The various OEM cruft and variety of optimization makes running application unpredictable (not even adding the variety of Android version in the wild if you want to cater for say 80% of the users).
So yes Android diversity might be one of its strong point, but makes designing application difficult.
Yes, as far as design goes, the original article is a complete clusterf**k: in his design specific points, he claims that the web has lost to Android but then claims the web as his proof for why generalized development is superior. If so, why is the web losing to app ecosystems? Do web developers actually think responsive design is easy or solved? (I don’t think so.) Do web developers think that responsive designs that target the LCD are just as valuable and just as easy to create as web sites that target a specific segment of the market? (No.) How many users curse being directed to a mobile web site rather than a full version? (Many.) Even if the responsive design problem were solved, haven’t we moved well away from the old Java model of apps that don’t exploit unique platform advantages in exchange for targeting the broadest possible market and moved on to specialized apps that leverage the unique pros and cons of specific platforms while using web services in the background to populate the specialized apps with data that can be shared across platforms? (Yes.)
Hell, this guy’s own product: Twitter and Twitter apps specific to these platforms in question show that he is wrong more than he is correct. (The initial explosion of Twitter apps on iOS, the backlash caused by Twitter killing off this app diversity, Twitter’s attempts to move Twitter forward on its apps first and the web second, that Twitter just design specialized apps with different features that are valued by the platforms, etc.)
Some will always argue that a write-once, run anywhere “responsive” design is superior; others will argue that a specialized design attuned to the specific features of the ecosystem will be superior. The truth is that both approaches will have their pros and cons. As soon as you lean to one more than the other, you are wrong and in need of greater clarity (something lacking in both of these posts).
However, I think from a user perspective, it’s more likely that users will prefer the specialized design (particularly if it is superior to an equivalent generalized app). From the developer perspective then, it’s a question of incentive. The original article seems to argue that ideology and altruism should override other incentives to please their potential users. I don’t find that argument very persuasive. The developer should be measuring their costs versus their returns on investing in a generalized design for multiple (or one, in the case of Android) ecosystems versus specialized designs for one or more platforms that may or may not have equal costs and most likely have different ROIs.
Edited 2014-03-24 19:21 UTC
As browsing statistics show, there is also a difference in usage.
Yes, there are more android devices.
Somehow, more iOS devices are seen by web servers.
This can only mean that an iOS device gets more usage (online usage as oposed to phone use) than an Android device.
Pair this with the fact that android owners using your app are less likely to pay for it, and you see why the money is still in iOS.
Reminds me of the days when a windows developer would do a lame job of porting an application over to the mac. The app would run but not be mac like, and would be awful to use. The mac only apps were usually better.
Of course with window they had to initially write there software and drivers to work on a much more diverse hardware set, so it was relatively easy to test a limited set of mac configurations.
But this would be like a mac developer porting to windows yet without the incentive.
As a longtime mac advocate, In the smartphone world I should become an android advocate based on some of the same principles , except that application quality has always been number one for me. The app must integrate perfectly with other apps. And be self explanatory. On the mac, all applications behaved the same way. On windows there were so many different application developement options that what you learned to do in one app just didn’t work in another.
Instead of suggesting that many iOS developers are lazy or unskilled, the author might want to consider the economics — specifically the return on investment (ROI) for developing on each platform.
In an interview with All Things Digital, Bob Bowman, CEO of Major League Baseball’s Advanced Media division, said that, “the Android user typically is less likely to buy, and therefore the ROI on developing for Android is different than it is for Apple.†Benoit Essiambre, creator of the “Speed Anatomy†app for both iOS and Android, broke down how his application sells on both platforms, noting that he receives roughly 50% less from the combined paid and ad supported versions of his app on Android compared to its sales on iTunes.
Along with less lucrative sales, there are higher development and support costs for Android, with the plethora of devices and so many active versions of the Android OS making it a developer’s minefield. If you’re going to see 50% of the revenue, why would you invest 200% of the effort?
This is the innovators dilemma.
…and the answer is that you (probably) won’t invest more into something that returns less.
Others will, perhaps those who have less to lose, and could be satisfied with those returns that are lower than what you enjoy.
…and in a few years they will be eating your lunch.
No, in a few years, they will be bankrupt. Getting less income for more investment is how companies fold, not how they succeed. You’re living an Android fantasy world if you think otherwise — it’s simple economics.
Developers are flocking to iOS because iOS development and support costs less, iOS users buy more apps, spend more money on the apps they buy, spend more on in-app-advertised products, and have more disposable income.
I don’t know about the disposable income part, as I am an Android user and have plenty of it, but this guy speaks the truth. What it boils down to is that, for whatever reason, iOS users spend more money, and is apparently easier to develop for, so that’s where devs give the majority of their attention.
On the other hand, Android is steadily making Apple a niche player in the mobile market, so in the next few years, the number of iOS users left may be too small to make a real difference anymore.
The number of iOS users continues to increase. They decrease as a percentage of smartphone users but actually increase as a percentage of all mobile users.
I’ve long argued that Apple has developed a sustainable ecosystem (an ecosystem that may not be larger than the entire PC and/or console market(s) yet but already as attractive). I don’t see an ecosystem selling 9 figures worth of devices per year (particularly when those 9 figures worth of devices require targeting just a handful of form factors, hardware specs, and software features and has a large number of benefits from a tight but feature-rich ecosystem) as becoming so unattractive that developers are pushed away from it. Certainly not in the “next few years”. Maybe if Apple had plateaued and then lost share in the US and Japan back in 2010 rather than growing to nearly half the market, but we’re long past that chance to truly marginalize the iOS ecosystem.
Edited 2014-03-24 18:04 UTC
No offense was intended. I’m speaking about published statistical averages, not you personally.
What matters is who spends money. Apple still sells about 21% of the smartphones in the U.S. and they hold a massive lead over Android among those with high incomes and/or advanced degrees.
Someone with a low-end, free-after-activation Android phone probably just wanted a free phone on which he can check e-mail, take pictures, and do light web browsing. He’s unlikely to spend a lot of money on applications.
Unless Google finds a way to reign in Android development and support costs, many developers will stick with Apple. Android vendors often let non-current phones languish with horribly outdated versions of the OS. Many phones that meet all of the hardware requirements for later versions of the Android OS cannot get those upgrades because the manufacturer has shelved all development and support efforts for the non-current phones.
Apple is much better about that, with iOS 7.x being backwards compatible to all phones back to the almost four year old iPhone 4. It was released in September of last year and, already, iOS 7.x is on well over 80% of iOS devices. By contrast, Android 4.x, first released in 2011, is on a smaller percentage of applicable devices than the half-year-old iOS 7.x.
“Apple still sells about 21% of the smartphones in the U.S. ”
Umm, you need to DOUBLE that number.
Edited 2014-03-24 18:48 UTC
I stand corrected! I must have inadvertently used the worldwide market share rather than the U.S. market share. In 2013, Apple sold 45% of the phones in the U.S. and Samsung lagged far behind with 26%. (source: NPD Group/Mobile Phone Track)
With Google rolling a lot of the APIs into Play services where Gingerbread is the base requirement, this isn’t as big of an issue as it used to be. (It also means new Android versions are getting to be quite boring, but I guess you can’t have your cake and eat it too.)
The percentage of Android which is not Google Play Services Android is growing, actually outgrowing Google Android, and is now approaching 50%. Attempts to minimize fragmentation by becoming more proprietary and closed have been significantly less effective at reigning in fragmentation than open source and cheap has been at increasing fragmentation in the developing world where the majority of Android growth is occurring… without Google services. KitKat growth in relation to just Google Android is almost inconsequential (relatively speaking); I’d wager it’s actually decreasing in relation to Google Android + non-Google Android. But, of course, Google and the many Android advocates aren’t doing anything to provide any insight to these factors; if anything, they are trying to obscure or ignore them. Which is unfortunate. The data is there, but we actually have more insight and analysis of the Android landscape coming through inference via the analysts who’d typically be considered in the “Apple camp” however objective they try to be.
Edited 2014-03-24 20:24 UTC
Two points here:
– The reason why vendors are able to rip out the Google services APIs and fragment the shit out of everything is BECAUSE it is open source. If it were proprietary, they wouldn’t be able to do that.
– If a device doesn’t have Play services, it probably doesn’t have the Play store, in which case it ceases to be Android anymore. Just like the Kindle Fire… these aren’t Android tablets, but Amazon tablets with a bastardized Android hybrid OS. People in 3rd world shitholes who are buying these devices probably wouldn’t have a means to buy from the Play store anyway, even if it were included.
Agree in both regards. Wasn’t this implied in my comment? Do you think this is a good thing?
IMO, the best approach would be a hybrid one. Make the license more restrictive, where people can still look at the source if they want, but forbid modifying and redistributing it, so companies can’t screw with it to the point where apps don’t work from one device to another.
On the other hand, make the stock rom open-ended enough so rom makers can put in the tweaks they want without having to resort to hacking the source code.
And that is taken from… a tramp’s posterior?
It’s taken from the same source as the initial source for this article: Horace Dediu.
https://twitter.com/asymco/status/447807943116083200/photo/1
Apple has sold 800 million iOS devices compared to 1.2 billion Android activations. Last year alone Apple sold 237 million iOS devices.
Last quarter Apple sold 870,000 iOS devices a day compared to Google’s 1.5 million daily activations.
That’s a damn big niche.
Edited 2014-03-24 22:09 UTC
Getting less profits does not make you lose money.
Who said you would realize any profits?
add to that iOS users enjoy their device more, enjoy their tech support, enjoy their purchase overall more, and also integrate that iPhone with several other apple products that exhibit the same ease of use and stability.
android is a knockoff and everyone knows that. it’s a good product and sure enough cheap androids have flooded the market, but most android’s aren’t “smart” phones because most of the users can’t find or won’t use it’s advanced features.
Far more interesting than this bit of third-hand advocacy playing off of Horace Dediu’s infographics is this one tweet in the midst of this small flurry of graphics/data:
https://twitter.com/asymco/status/447992531213885440/photo/1
According to Horace’s projections and numbers, Samsung may account for more than 70% of non-AOSP/Google Android. Horace himself seems to be questioning that finding, but if it’s remotely accurate (and that seems likely), there are far more interesting conclusions there than in the data that Horace has been providing at regular intervals for more than 5 years now.
I’m having difficulty finding any value in either Thom’s or the original post. Firstly, they are actually saying opposite things: the original argues for generalized development and addressing the largest possible market for seemingly altruistic means and “world enrichment” (this is very confused, he wants to argue that the larger market is more desirable but doesn’t want to hear arguments that there is less value in that market actually making it less desirable but he does say that someday the revenue will be there, but it’s irrelevant (?!)). Thom actually argues for specialized development for economic reasons but he doesn’t point to any actual economic reason (or at least he doesn’t seem to be arguing for it on altruistic reasons), rather, he seems to be blaming generalized development or iOS-first development for the lack of an economic incentive.
I’m completely lost on the original’s argument that the market is large thus worth developing for, but any argument trying to formulate that value can be dismissed because it’s “economic prejudice.” And then he argues that someday in the future it may actually be economically prudent (but I thought economics shouldn’t matter? But why should size of market matter then either? Ideology only? Maybe that’s your answer for why it is failing on actual economic and reality-based metrics?)
But there are just some insane, weak arguments. As already mentioned (above), he clearly and specifically glossed over Horace’s graph that gives almost 50% of the Android market to non-Google Play/Services devices. He specifically mentions KitKat but ignores that it is less than 10% of the install base.
If anything (and I really don’t see much value in either of these posts), both of these posts seem to be a good starting point to clarify why Google Android is an UNattractive development platform.
Edited 2014-03-24 18:52 UTC
Horace Dediu makes no such statement. AOSP and Google Play Services are not the same. Nexus 5 is non AOSP, only some Play editions of devices are AOSP. In fact, most Android devices are not AOSP. And not surprisingly, most of the mare Samsungs that all have Goole Play servies…
Oh no…?
https://twitter.com/asymco/status/447807943116083200/photo/1
Doh!
As an Android user and developer (and iOS user) myself, I would see it as follows:
1) Many fine applications on iOS just have a bad port on Android. Simply take a look at two of the most popular messaging apps, WhatsApp and LINE, and you would know. Design of WhatsApp is just brilliant. Not only does it adhere to the Android design guideline, it looks great on it too. On the other hand, LINE is just a port of their iOS version, even compiled with an Android 2.x SDK. So it looks really bad on Android.
2) For version fragmentation, frankly speaking it has nearly nothing to do with Android design. As developers we are talking about API levels. We concern on programming APIs that only exist on, say, Android 4.x or above, and those APIs won’t work on Android 2.x/3.x. However in terms of design, there are still a lot of flexibility. Even for the holo theme or the Action Bar, there are alternatives (ActionBarSherlock anyone?) that works on Android 2.x.
Look at those popular apps like Facebook, WhatsApp, or Twitter, or all the Google apps. Do they require Android 4.x to look good? NO. They look good on Android 2.x too, and the interface is unified across different Android versions. These are the design efforts that these apps devoted and as a result, they look great.
I wholeheartedly agree that most poor designs on Android are simple laziness, or should I put it in other words, “minimize cost, maximize profit”?
A good developer is a good developer no matter what he codes for. A developer who is good only at one thing isn’t a good developer at all.
A specialist tends to perform his trick highly efficient. Yes, a specialist requires a lead. Some one who tells what to do. Some one who translates customers’ slang into the slang of the specialists’ domain.
But why should a specialist be considered a bad developer?
“Inquiring minds want to know.” [Soulbender]
Greetings,
pica
Edited 2014-03-26 08:25 UTC
There may be millions of Android users, but the fact remains that they spend a paltry amount of money on apps and device-related purchases than iOS users.
There may be millions more Android users, and by definition more Android programmers, and there may be some beautiful Android apps, but the fact also remains that with all those apps and programmers there are hundreds of Android devices, a huge amount of which, due to their being sold with, and never upgraded from, an ancient version of the Android OS, cannot take advantage or display to those users all the beauty and capabilities of those apps.
There are a lot of generalizations and rationalizations in Thom’s commentary on the referenced article, but few of them have any meaning at all to the above factors.
I primarily develop for Android, and try very hard to make sure to follow Android user guidelines and Android style, still the iOS versions of my apps outsells the android version by 3:1. This in spite of the fact that I don’t regularly use iOS devices and that probably breaks a lot of UI expectations of my iOS users. So my conclusion is that the lower sales on Android are not related to poor quality.
I would rather suspect that it has to do with higher availability of good free apps for Android. Partly Google is to blame for this, as it took a very long time before it was possible to charge for apps on Google Play in many countries, resulting in the market being flooded with good free apps.
my stereotyping of phone buyers (in the US):
A) new iPhone with a bright shiny case = fashionable buyer, spends on games and fluff, talks to Siri alot, doesn’t know or care much about the OS or anything technical like that. Might not even own a computer. CC# is on file with AppleStore and iTunesStore and purchases occasionally.
B) old iPhone loaded with miles on it = heavy business user, spends on apps & accessories, knows the OS pretty well, and would buy another iPhone 20 minutes before their old one died. CC# is on file and they purchase often. Apple might get 2-10 app/media purchases a month from this user.
c) new Android with huge screen = fashionable budget buyer, plays free games and does social media, loses most of their on-phone creations, has a new “better” droid headset 4 months later with a even more retarded looking screen and more software confusion. Probably doesn’t own a computer, or at least doesn’t sync that phone with their databases.
D) stock Android + a nasty look towards any iPhone user = geek android buyer, who got a nice phone for a nice price and is determined to show that they can be down with the jones’ trendy lame iPhone. they use words like fanboy and cult to explain why they didn’t purchase the more expensive and more respected iPhone. there’s usually some new/obscure tech they are preaching about that isn’t available on the iPhone (or not default). they also live on OSNews ;-).
E) windows phone? – I don’t think I’ve seen one in the wild since about 2008. These people must really have major investment in Microsoft, both in money/licenses and head space. Microsoft has been looking behind the times going on 10+ years now and the marketplace has responded. To this day I see more blackberry’s in use than winphone. I did see 1 surface rt at a local cafe last year but haven’t seen it back since.
F) old-school phone — someone who tried an android and now can’t stand smartphones, and just wants to make calls and text.
ok, end of phone stereotyping. i’m in the B) category right now, with my loaded iPhone 4 starting to really drag ass after 3 years of living on it. I’ve had a smartphone in pocket going on 17 years now, damn!
The relationship between user base and developer needs that Thom drew is spectacular. Ten minutes in, I’m still laughing.