Is Choice a Problem For Android? 361
New submitter mjone13 writes "Dave Feldman, in a blog posts, says that the problem Android faces is giving consumers too much choice. He cites several studies which state that consumers generally are unhappier when they have too much choice. 'Catering to all individual preferences creates a bloated, bland product. Not to mention a UI that’s impossible to navigate. Furthermore, people are notoriously bad at identifying what we want. And what we do want is influenced heavily by what we know — our expectations are constrained by our experience.' He then goes on to talk about Android fragmentation, app developer problems and bug issues. Finally he says the people who general prefer the choice Android provides are tinkers similar to gear heads who love tinkering with their car. 'I think many who extol Android’s flexibility fall into the tinkerer category, including some tech bloggers. They love all the ways they can customize their phones, not because they’re seeking some perfect setup, but because they can swap in a new launcher every week. That’s fun for them; but they’ve made the mistake of not understanding how their motivation differs from the rest of us.' Is choice really a problem for Android?"
Whether it's a problem depends on what the goals are. Providing a satisfying experience to a bunch of tinkerers is a very different thing from providing a satisfying experience to the multitude of non-tinkerers who buy smartphones.
Nope (Score:3, Interesting)
I'd rather have too much choice instead of barely any.
I'm well aware of that famous TED talk where the presented talked about the paradox of choice (http://www.ted.com/talks/barry_schwartz_on_the_paradox_of_choice.html), and it does have some merit. But the way I see it, appealing to the masses means that those who don't fit the mold are generally left in the dust because it's not "economically viable" to cater to them. Keeping things open and keeping choice available means that there's something for everyone.
Choice requires a greater level of personal responsibility. You can't (and shouldn't) rely on some corporation to make all your decisions for you without being able to change them if they aren't suitable for you. It might be easier to just go with a monoculture of decision-making, but you'll pay for it once you realize that you aren't like everyone else.
Re:choice doesn't *require* bad defaults (Score:1, Interesting)
Maturity required (Score:4, Interesting)
I agree that there is simply too much "crap" in the Android markets all over. The amount of good, quality, useful stuff is a seemingly small ratio of what's out there. But I get by just fine and enjoy a good quality experience. How?
Just like with Windows computers and the like, you simply have to limit what you do with your machines. Limit the resource consumption. Limit the amount of apps you run. Limit tweaks and [animated] wallpapers and all that junk. Do the things which are useful and stop trying to entertain yourself with a new toy every 10 minutes. I take advantage of the fact that people out there are dumb enough to try every app available. I get to read reviews and comments to assist in the choices I make. Good for me, bad for them when things don't work out.
Maturity is required. The market of available crap is not to blame for consumer behavior. (This statement is in sharp contrast to my position on the food we have available to us in the US... the market *IS* to blame and especially when they fight consumer choice and knowledge by preventing information from being available to consumers so they can make their own decisions.) The users are making all the choices... and they always will.
Make good, informed choices. Give favor to software makers with good reputations. It's not that hard.
Re:choice doesn't *require* bad defaults (Score:3, Interesting)
Why is everyone talking like there even is a problem? In August Android had almost 80% of the market [techcrunch.com]. Yeah, it must be incredibly boring and horrible to use if so many people want it.
Yet Apple's profit was greater than the other's combined; with Samsung a close second in terms of profitability. It's really just a two horse race with Apple vs. Samsung; so in that sense as long as Samsung maintains a consistent UI and feature set the "too many choices" argument is not relevant. It's shaping up a lot like the PC market did - Apple has it's own proprietary offering; the PC has MS-DOS but each implementation is customized often to the point of a lack of compatibility until IBM essentially set a hardware standard that allowed the emergence of a standard OS implementation. The biggest difference, other than the much greater difficulty difficulty in making Samsung clones like the old IBM clones, is Google didn't chose to make money on the OS like MS; it remains to be seen if that will ultimately was a good choice.
Re:choice doesn't *require* bad defaults (Score:4, Interesting)
As far as I know Android has a standard as well. You use Java for hardware compatibility and C++ for speed. Most apps run on all modern versions of Android devices. MIPS processors are the only exception that I can think of. Does every app run on every iOS device? No, because of difference between the devices.
The best analogy I can come up with are trading card games.
* The card game Apple only sells their game in pre-built decks. Everything is same so everyone can play the same game. If you change the deck the wrong way (mods and unofficial add-ons) you may not play with the other kids.
* The card game Android sells everything in booster packs. You get more variety and can pick up cheap decks of discarded cards. But the cheap decks may have old cards from previous game version that are not compatible with way most people play the game. Or you have someone who decided to make their own cards which don't completely follow the core rules. More fun to play for certain people but doesn't work with the game as a whole (we're looking at you Kindle).
Which one would you have more fun with for the price?
Testing for 123 versions is a real pain (Score:4, Interesting)
Went to a presentation on a project that's released its web tool as an app (iSpot [ispot.org.uk] - a nature spotting community tool). the project leaders said that at the point they decided to develop an Android app version, they asked the technical team to identify how many different versions/configurations of Android were out there that they'd need to make sure the code presented well on, to ensure a good user experience for all (you really don't need your first reviews on Google Play to say it sucks on their device in their preferred configuration). Apparently the technical team identified 123 versions/configurations of Android (approximately early 2012).
The project leader said this makes it a nightmare to test for a small development team (about 4 employees on the project). I am not sure what the solution is but it sounds like it causes them a lot of pain and requires a lot of management to ensure the majority of users get an equitable and positive experience of the app.
Re:choice doesn't *require* bad defaults (Score:5, Interesting)
Not to mention all of the different versions, you have people running everything from 2.1 to the latest version. For our app on android 63% are running 4.0.3+, however 31% are running 2.3 so we can't abandon 2.x yet. With iOS you generally only care about previous version. So currently target for 6.1 or higher, if you REALLY want to squeeze out every last download stat you can, aim for 5.1 (since ipad 1 can't upgrade to ios 6)
http://david-smith.org/iosversionstats/ I find these statistics fairly accurate. Since we don't use any ios6 features we target 5.1+.
One of my cousins is a product manager at a medium sized mobile gaming company, he said they have a similar experience to ours. Their games download stats are about 10-20% android rest iOS, yet they spend 3x the resources on android support as they have about 90 devices laying around to test on, and people always calling up because "it runs slow on this device", "it crashes on this device". They completely killed android development except for their top couple apps
Re:choice doesn't *require* bad defaults (Score:4, Interesting)
I do know it isnt necessary for android to test against every device, like you we only test against maybe 2 devices and then let crash reports catch the rest, but the issue is if there is a crash report for a device, and the crash isn't painfully obvious and research doesn't turn up what the issue is, then you might be forced to go out and pick up the device.
Re:choice doesn't *require* bad defaults (Score:3, Interesting)
Apple's target consumer is the same consumer who pays more to buy the brand name milk cause they think it tastes better. Having worked as a teenager in a milk plant, I can tell you that the only difference is that they change the label on the carton.
Re:choice doesn't *require* bad defaults (Score:4, Interesting)
Bullshit. I buy apple devices because if/when they break, i walk into an apple store, say "this is broken" and they give me a new device and i get on with my day. I don't get the "make the user play troubleshooting technician" bullshit, i hand the broken device over, and they hand me a working one.
Customer service like that IS NOT FREE. Yes, the devices cost more for lesser on paper spec. I don't care. I'm not paying for some on paper spec, I'm paying for a supported device.
Re:choice doesn't *require* bad defaults (Score:5, Interesting)
A lot depends on who and what you are targeting. If eyeballs and advertising is what you after supporting Android is a must. But if you are after in-app purchases as your revenue model, it's iOS you want.
I've been developing mobile apps since 2009. Early on I was making as much off ad revenue vs app purchases, but by last year the ad revenue went in the crapper. So much so that I stopped releasing updates for android. By that time Android accounted for a little over 60% of the installs. It accounted for less than 15% of my revenues. Android accounted for over 90% of my complaints and requests for support because someone with a cheap pay-as-you-go android phone would run into a problem on a device I didn't even know existed. I was making at most a couple thousand a month from the apps, mostly from iOS users. It was enough that it paid my basic living expenses like rent & utilities meaning my day job work could go into savings. But it wasn't enough for me to go out and buy every freaking handset on the market at $600 a pop.
Now on the professional day job part of the world we usually price for iOS first and includes QA for current generation and usually the previous 2 generations before that. Right now if you paid us to write an app, we'd ensure compatibility with the iPhone 4, 4s, 5 & iPad 2, Retina, Mini. Next month it will likely be 5S/C, 5, 4S & iPad Retina, Mini, + whatever is announced next week.
For Android we will test against Nexus Phone & Tablet and certify QA with those devices only and it costs our clients about 1.5xiOS. Why? Because we know we'll be answering "QA for XYZ handset was not covered in the contract" a few times. So we build it into the price of the contract. We do offer QA for additional handsets & tablets @ $5,000 per Android handset/tablet. Most of our clients will maybe ask for QA against the latest Samsung Galaxy devices and that's it. Only one that I can think of asked for Samsung & Motorola because the boss man had a motorola phone.
When Android first started we tried to QA against as many handsets as we could and we were losing money on those contracts. When Google released their official devices we decided, even though nobody used them in the mass market, those would be what we'd test against. That was the "official" devices for compatibility. What handset makers & carriers did beyond that we'd have to charge extra to fix because we'd run into the same model android phone would have odd quirks between different carriers sometimes.
just like windows had 90% of the market? (Score:4, Interesting)
and MSFT was thinking, why are we even talking about problems, we have 90% of the market. until they didn't.
marketshare size doesn't mean everything is dandy.