Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Google's mismanagement of the Android Market (nanocr.eu)
96 points by mbateman on June 28, 2010 | hide | past | favorite | 64 comments


The Android market does not make me want to buy apps. This is its greatest failing.

Apple's app store actually feels like a store. It's fun to spend money there. There are many different ways to discover new apps of all types. Each one typically has as much info and screenshots as I want to see.

In the Android Market, I have the category charts that never change, the firehouse of junk and spam that is "just in", or search, which is only really useful if you already know the name of the app you want. The descriptions are tiny, the screenshots are ugly and often missing, and the comments are mostly YouTube caliber.

If they can't make it fun to buy Android apps then the whole platform is going nowhere.


I don't believe I've ever had fun buying PC software, but that platform went somewhere. What is your reason for believing the smartphone market is different?


"somewhere" being primarily Microsoft, Adobe, and virus scanning applications, all of the above for large amounts of money and frequently purchased as necessities. The market for smaller desktop applications is pretty hard, and seems to mostly get along by the PC user-base being utterly massive, not by any ease-of-use.

Applications on OSX are a bit nicer, as there's rarely an install/uninstall process at all, and an enormous amount of the people I know with macs have bought several smaller applications for various minor needs. Nothing that can't be faked by text files / Excel, but a nicer interface makes a difference. Things like Sparkle also mean people update far more frequently. I don't see that much in my Windows-using friends, especially in the non-geek sectors. But OSX has a much smaller market share, and I have a feeling that a lot of the successes are due to things like MacHeist, pulling in $1/2 million+ over a couple days, which do make buying software fun. One can argue that they also dilute the money pool by devaluing them when not on sale, but the amount of effect that has is nebulous at best.


Because there is competition in the smartphone market, and someone has already raised the ante from "not fun" to "fun".


A few other issues:

-- There's no in-app upgrade option. If you install a lite or trial edition you have to search the Market for the full version and install it on top of your lite/trial version. So if you happen to search for an app, buy it, and install it sometimes it just won't work (because you were supposed to install the lite/trial first)

-- Apps without meaningful descriptions should be deleted or hidden by default. A changelog is not a description.

-- Apps that start off saying "doesn't work with..." should be deleted or hidden by default or only shown on the Android devices it does work with.

-- It would be fantastic if you could sort searches by downloads, ratings, etc. You'd think since Google is a search company they could figure this one out. Maybe in 2.2?


Most of the problems you describe could probably be policed by Google, but Google has asked the developers to do this, and it is definitely a problem.

POINT 1 -- There's no in-app upgrade option.

True, but you can quite easily link RIGHT TO the app you require your user to purchase. It's actually super easy to do, I don't know why some don't do this. However, maybe you are referring to the ability for an app to initiate a self-update, in which case, I'd love to know how to do this.

POINT 2 -- Apps without meaningful descriptions should be deleted or hidden by default. A changelog is not a description.

App developers are given 325 characters to tell the entire story of their apps AND describe why we require an update. It's awful. Have you ever written a Haiku? Every time I try to create an app description AND include a changelog, it's like geeky Haiku.

I feel like an idiot replacing the word "for" with a "4" because of this space limitation. It's embarrassing.

POINT 3 -- Apps that start off saying "doesn't work with..." should be deleted or hidden by default or only shown on the Android devices it does work with.

App developers can put a target SDK, minimum SDK and even max SDK in their Manifest. I'm not sure why more don't.

-- It would be fantastic if you could sort searches by downloads, ratings, etc. You'd think since Google is a search company they could figure this one out. Maybe in 2.2?

Agreed.

Though I'd like to one-up this and ask that there be a "Featured Apps" option where Google employees show off obscure but cool apps.

Then, how about a "Sports" category in Games? I for one would be really happy to see this.


the ability for an app to initiate a self-update, in which case, I'd love to know how to do this.

Seen Sparkle[1]? It's all the rage in the OSX world. Thousands of applications use it. It'd be a good place to start. Beyond that, there's also Google's Update Engine[2], also for OSX, which does updates file by file, and I believe can even work with just changesets (Sparkle replaces the whole .app bundle). At least, I think Update Engine was started by Google...

Or, if you want to go for "epic" in-place self-updating, something could be done which mimics how some Linux kernel upgrades can be swapped while the OS is running. That's a neat trick. I believe Nginx can do the same, it'd probably be an easier start. Unless the binary is locked when downloaded, I don't see why that at least shouldn't be possible.

Though, all that said, IANA Android developer, and don't know what limitations are in place. Could be it's just flat out impossible, I honestly don't know.

[1]: http://sparkle.andymatuschak.org/ [2]: http://code.google.com/p/update-engine/


> Seen Sparkle[1]? It's all the rage in the OSX world.

Sparkle doesn't exist to upgrade your app from e.g. paid to free.


Minor nitpick, compared to it not being open source, so it's not an ideal one to begin with. There's no reason a nearly-identical system couldn't be used as a paid-upgrade tool; just download the license file that was purchased in-app, and have the newly-updated application read it in.

It was chosen for conceptual design of something which has radically changed the application-updating playing field by being so successful. Normal people update regularly with Sparkle; how many have you seen ignore a Java / Windows / Adobe update for the thousandth time?


> Minor nitpick

Uh no it's pretty major.

> compared to it not being open source, so it's not an ideal one to begin with

Sparkle is not open source? Are you high? http://github.com/andymatuschak/Sparkle


gah, my mistake. Must've been thinking about something else.


"There's no in-app upgrade option."

It's possible to add a link in your app to the full version, such that when it's clicked, the market page will open and the user can just click 'purchase'. It's not exactly in-app, but it works pretty well.


Apps without meaningful descriptions should be deleted or hidden by default. A changelog is not a description.

Related to that, the 325 character limit for app descriptions is absurd.


I see a lot of developers complaining about this, but from the point of view of a user, that feels like the limit of how much users (me) would be willing to read per-app.

In my experience, it's rare that an app cannot be adequately described in that space.


Your comment contains 257 characters. Do you seriously think adding 72 more would allow you to effectively describe an app of any significant complexity?


It's difficult, but it's mostly doable, and more importantly, necessary. People don't read, and even if you had a longer description people aren't going to read more than a few lines of it. So instead of losing out because people didn't read the whole of your description, I see this as an incentive to place everything within the region people will read.

However, I do agree that with apps of increasing complexity (WaveSecure comes to mind), you'll have to leave out a lot to fit in the space. In these cases though, I feel the rating and comments should give the user enough push to try the app, after which your app can describe itself all it wants.


One weakness of the Android market that isn't discussed as frequently as it should be is that most [1] developers can't actually sell apps for Android.

Unless you live in one of the countries listed on http://www.google.com/support/androidmarket/bin/answer.py?an... you can't sell paid apps.

Too bad for all those Indian & Chinese developers who've built out the Apple AppStore's inventory. On Android those positions in the catalog get filled up with free, half finished apps that the developer abandoned when they realized they couldn't sell it.

And those free apps remove opportunities for paid apps to develop.

[1] Yes - I understand a lot of apps come from countries on the list. I still argue it excludes more than it includes.

[edit: corrected link to country list]


yeah this really sucks. Just got my first Android Phone, and am keen on getting stuck into developing for it, but I am Australian, so I'll have to go around the long way. If Google wants to catch up to iPhone in terms of apps available, it should be putting more emphasis on this.


Sidetracking a bit, is it possible for developers from those countries to display ads and raise revenue from them? It appears as though Google is encouraging that as the primary business avenue on the Android, and a lot of Android app developers make equal (if not more) money from ads as from sales.


Wow, I didn't even know they don't cover the whole of the EU countries -- disappointing.


It doesn't even cover Canada, which is a close geographic and legal neighbor of Google.


Yeah, this simplifies matters quite a bit.


So, the moral of the story:

Google needs to make their Android store more like Apple's app store. The openness is getting filled by crap / spam / questionably-legal apps.

Someone actually admitting that there's a benefit to filtering! Who saw that coming?


Android users and developers aren't totally dependent on their app store like iOS users are since they can install software from outside the store. This is where the openness comes in, NOT because of what they chose to allow in their app store. For this reason Google could actually make their app store a lot stricter than Apple's without attracting the same level of criticism. However I think that both companies are keen to see the 'total apps available' number increase regardless of quality.


But how many non-geeks will be installing software outside that store? The first doorway matters, and it's been left wide open. There's even a note pinned to the frame inviting vampires across the threshold.

Personally, I'd think a multiple-store method should work best, where each store can control what's in it. A "kid-safe, parent-approved" store would be useful, as would "experimental", "anything goes", and "cryptic dogma but generally good". A way to select which ones you want to look in, and you're golden.

Though I realize I've just described a package manager with configurable sources.

(unless this is possible currently? haven't heard anything, and it would make the blog post nearly a moot point...)

--- edit:

I was referring to the openness of the store, not the platform, as evidenced by referring to "[Google's] Android store". Maybe it would've been clearer if I said "Android Marketplace"? Should I edit it while there's time? The openness of the platform is utterly fantastic, I wouldn't want them to change that for anything.


> But how many non-geeks will be installing software outside that store?

Depends how strict the app store policy is I guess. If Apple allowed this option then I'm sure a lot of geek and non-geek users would take up the option.

> Personally, I'd think a multiple-store method should work best

Agreed. Payments would be difficult and I guess they would want to discourage a 'pirate store' as well. But different listing criteria would be useful.

> I was referring to the openness of the store, not the platform

>> Someone actually admitting that there's a benefit to filtering! Who saw that coming?

I don't think the idea that assisting users to filter their options was ever controversial or surprising. With your multiple-store idea the openness of any particular store wouldn't be such an issue.


There's been a redonkulous amount of uproar over Apple's filtering of their own store. Often coinciding with people saying the world would be a better place if only Apple would allow their app, and if puppies fell from the sky, and nobody had to go to work, they could just bang on these drums all day.

Granted, it's the only store, so it's a different situation. It's just a nice reversal, where Android fanbois have been ripping on Apple fanbois for having such a closed store, and now the benefit of that closing is being made apparent. There's not much Android ripping here, most people here are at least somewhat geeky, and would love to see Android become bigger and better than iWhatever.

edit: as to the first point, I'll still say "very few", unless drastic steps are made to make it easy to choose from the very first interaction. There's a lot of change-fear out there in normal people - change means you don't know where your favorite button is any more, and it now breaks in new and exciting ways instead of the ones they could somewhat predict.


I'm not really that interested in fanboyism. The principles at stake in comparing open vs closed systems are orders of magnitude more important.


That's true but he is still correct. While Android can (and should) keep the ability to download and install "apps" (I am surmising that an app is some kind of packaged binary or deliverable like a .tar.gz), the need to make their App Store more like Apple's App Store is critical. They need to have strict guidelines and restrictions and say "you can't put this crap on our store." The difference is that if the dev disagrees with Google they can go post their app on their website. Is this Google controlling their ecosystem? Yes -- and rightly so. Google's App Store has their name on it for a reason and the apps accepted into it should reflect at least some of the professionalism displayed by the company itself.

Of course, the problem with this is that Google is completely awful at product management. As far as engineering goes they have a great deal of success but I don't think anyone would deny that their customer service is abysmal.


exactly. Google always has great engineering, but poor customer service. I dread the day I can't get into my Gmail account...

They work as autonomously as possible, and don't like stepping on peoples toes. The Marketplace needs some strictness, if you want to download dodgey ringtone apps, you can go to the browser and get the install from there. But the marketplace should be far better than it is now. Not even just talking about the fact there is dodgey apps there, the search function isn't the best... THE SEARCH FUNCTION GOOGLE! I tend to go to those app review sites and use the bar code scanner, which I shouldn't have to do to find apps from the worlds best search company.


Of course there are benefits to filtering. There are however many different ways to apply filtering. Apple's app store and the Debian Stable repo are both filtered and curated, but no one would claim they are similar in either intent or approach. I want the App store to be filtered, but I want it to be filtered more like the Debian repos are filtered and, more importantly, for it to be just as easy to bypass the filter.


The moral of the story is that the author wants it to be more like apple store.

I don't and do believe that the fact that he wants makes him worse than a retarded dingo.

His only valid complaint is that google can't seem to get their ass out of their own payment system, which is indeed a tragedy.


I'm sorry, but how many fart apps are there in the iStore? How many packaged (and obviously copyright) web image viewers of various models?

Google's system has problems. Apple's system has problems. Apple's curation has almost entirely been focused on eliminating duplication (even if better) of Apple-provided solutions.


Apple's curation has almost entirely been focused on eliminating duplication (even if better) of Apple-provided solutions.

Do you have any sources to back that assertion up? In the WWDC keynote, Jobs indicated that the three major reasons for rejecting apps were bugs, use of private APIs, and apps not delivering what they promised, which would be a direct contradiction of your claim.


how many... how many...

Fewer. Comparatively, at the very least.


There are hundreds of fart apps in the AppStore. Hundreds.


Out of a quarter million[1]? Drops in the bucket. The article here is showing piracy apps with 250k downloads in the top apps list. A quick look at their top free multimedia list shows several ringtone clones, clearly implying many below that.

[1]: think I recall seeing that number recently, though I don't remember where. Please, correct me if I'm wrong.


Have a filter option in the Market app a little bit like safesearch. You'd have a single slider. 3 or 5 positions (low moderate high etc).

Filtering would work in a metascore comprised of several metrics. The metrics would be opaque and Google would tune them regularly like it does with it's search ranking.

Example metrics might be:

1. 'Spam' score based on app description.

2. 'Spammy Developer' score based quite simply on the number of apps released by this developer (this would immediately clean up most dross but people would quickly find ways to game it)

3. User ratings (weighted by the rating given to that user by other users)

4. Ratings derived from Google's web rankings of the app's web presence

Problem solved. Can I have some money for that please Google?


I may be misunderstanding your suggestion, but this system would require Google to admit that there is spam in the store, no? And if there is spam (as we can clearly see) why not just remove it instead of asking the user to manually hide it?


Becuase they would then be committing to a manual review process and all the scaling problems that would incur.


Isn't this just an opportunity for somebody to start their own android store and police it better?


Also, if you don't like the browsing experience in the Market application, there are alternatives for that too. Like AppBrain, which lets you browse for things online, organize what you find into lists, and queue applications for install on the phone.


Kind of reminds me of Google's mismanagement of Youtube. That's full of copyright infringement and yet folk seem quite happy to put their stuff on there.

I doubt they would have grown so big, or been able to win their recent lawsuit, if they'd had to green light every item that went up in advance.


The problem is not only the existence of such Apps on the store but that they rank so high. Google should know how to give good apps a prominent spot and rank scam apps low. That would solve many problems without censoring or banning anything.


Judging from the people trying to buy an Evo on launch day, it wouldn't surprise me to find that the apps listed are the most popular.

Some of the best apps simply aren't popular. Locale is an essential, but it's $9.99 price point (!!!) makes it unattractive to all but the most hardcore users. Free ringtones may be useless, but people like free, even if it's shitty.


If that's the case, maybe Google should "feature" apps based on more metrics than simply popularity, if they don't already. It would be neat to see some sort of app suggestion engine where Google shows you personalized app results based on your download history.

Maybe I'm optimistic, but I think we're going to see a lot of quality-of-life (and fragmentation) issues ironed out once Android hits a more stable release cycle where development is more focused on optimization and bugfixes than adding new features.


I've probably reported about 15 scam or infringement apps in the market to Google.

Zero have been removed.


You know, despite the Android app store having plenty of these crappy "Free ringtone!" apps, I'd still take it over Apple's controlled/censored (yes, I would call some instances of their app removals censorship) app store. It doesn't really matter though, since I can install whatever I want through .apk anyway, but I'd just like to say that not all users are getting fed up and asking Google to remove apps.

That being said, damn browsing Google's App Store sucks. It's horrible. Seriously, there needs to be some much better organization, lengthened app descriptions, easier to use and better filters, and just an overall better experience.


Given the openness of android, can we make our own 'better android market' application + website? May not be worth the time investment given most people would use the market anyway, but might make Google wake up...


Another point is that the Market application is Google proprietary - when you grab the AOSP source, you don't get the Google apps (Maps, Market, etc.)

Android is open source and manufacturers can use it without the consent of Google, but to include the goodies above they have to go through a vetting process (not saying that's a bad thing).

A completely open source implementation / replacement of the Google apps would be very useful.


I had somebody steal my content and create an iPhone app from it, too. The copyright holders should contact the Google Market about the infringements.

That said, the market certainly is not perfect yet. But not too bad, either.


Agreed. The Android Market is an absolute ghetto compared to Apple's App Store. It makes me think there's room for a third party to build a better, more heavily curated marketplace for Android apps.


Why in the wide, wide world of sports does a "iTunes App Store" app even exist in Android's storefront? Who would download such a thing, even assuming it's some kind ad-impression scam app?


Someone who is interested in running the cool new app they saw on someone's iPhone. Most consumers aren't nearly as interested in learning the ins-and-outs of what Android is versus iOS as you and I are.


The whole iphone/ipad business works because there are people who don't care about openness, things which look obvious to me and you are not as obvious. Many businesses work just by exploiting the ignorance of buyers. Apple is master of that art.


Apple grew to become the 2nd largest US company by market cap through customer ignorance? Yeah, that sounds plausible.


hence the raving over how non-category-defining the hardware is.


Its not the ignorance of users by company, its the ignorance of mass about the open standards we talk about. Being big doesnt mean you are doing it all good.


The proof is in the pudding.

Whether you like apple or not and there is certainly much to not like about them even if you like most of their products, you can't ignore that they have built the strongest digital ecosystem out there. Stronger than FaceBook or any other competitor.

Contrary to most others this ecosystem is built on paying customers which is quite an accomplishment. It includes both music, video, applications and now the Apple Store app which gives you access to the apple products. I don't even want to know how many people will be tempted to press buy on that one.

This is not about ignorance of mass, it's about reducing the friction to almost zero while still maintaining some level of quality.

In all other business situations this is the main goal.

What really baffles me about the Android App store is just how much friction there is. It's a complete mess to be quite honest.

Even though google should be applauded for their open approach, let's not forget that they do this for one thing and one thing only. To expand the reach of their adds.


"ads" not "adds" sorry


To reply to the first paragraph of the article-

>Earlier this week, CNET ran an article critical of the permission model of the Android Market. Google’s response to the criticism was that “each Android app must get users’ permission to access sensitive information”. While this is technically true, one should not need a PhD in Computer Science to use a smartphone. How is a consumer supposed to know exactly what the permission “act as an account authenticator” means? The CNET opinion piece “Is Google far too much in love with engineering?” is quite relevant here.

Is it?

Just to be clear, it is my belief that on the iPhone all apps have the run of every bit of functionality provided by the sandbox. Meaning if one app can access contacts, all apps can access contacts. If one app can get your location, all apps can get your location. If one app can pull your contacts and your location and send it to a web service, then all apps can do the same.

On Android this isn't the case at all. On install you had to explicitly confirm the required rights.

The Android model is a thousand times better, at least assuming they avoid user fatigue where people simply OK anything.

It could see improvements for sure, though. For one I'd like to see optional permissions. When I installed Barcode Scanner I was a bit perplexed as to why it asked to access my contact list. Turns out it's because it can create barcodes from contacts. Nonetheless, that should be an optional permission that I could set to yes (checked), no (red x), or ask on demand (cleared), defaulting to ask on demand. It would make me far less nervous about apps that seem to unnecessarily ask for the kitchen sink.


This is mostly true, except for location. The iPhone has always asked for permission before providing your location to apps (even to Apple's preinstalled apps, like Maps) and as of iOS 4, you're able to see a list of apps that have asked for your location, whether you said yes or not (with the option to change) and whether the app has accessed your location in the last 24 hours.


the link doesn't work for me

Edit: it works now


The Internet has child porn! Let's all switch to AOL!




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: