Android is more and more like the mobile equivalent of desktop Windows. Now they have the huge pirate market in China too.
When this happened to Microsoft, certain strategy-oriented parts of the company were happy to see Windows spread everywhere as quickly as possible and deal with piracy later, whereas others were more concerned with the bottom line and how to extract revenue immediately. Google will have to deal with the same political questions.
It's funny how Android even has its own versions of OS/2 in the form of BlackBerry 10 and Jolla's Sailfish -- that is, competing operating systems that proclaim to be technically more advanced while offering reasonable app-level compatibility with the dominant system. (I suspect that BB10 and Sailfish will share OS/2's fate in a few years. Too little user benefit against the high cost of lacking access to the dominant vendor's services.)
Its worst than that. A least with Windows on PC clones, you have control over the hardware and can update to your hearts content. Android makers often change the interface and can interfere with you from getting updates.
That's probably something that the PC clone vendors of the early Windows era (1990-95) would have loved to do, but they just didn't have the software competence.
Developing a new shell for Windows was technically possible, but must have seemed like an outlandish prospect to companies like Compaq or Dell. ("Where are we going to hire a bunch of C developers who know GUIs and can do a better job at this than Microsoft? Forget about it.")
Some companies did go there -- for instance, HP had an advanced Windows shell called NewWave [0], but I don't think they ever built PCs specifically for it.
Today, the situation is very different. The executives at any old phone maker will look at Android and say: "That's just Java, isn't it? We have plenty of people who know Java. Hell, there's a school down the road where they train 100 Java developers every year. Let's do it! We can be a software company too, just like Apple!"
And of course we know how that turns out. But it will take a long time until the executives at all those LGs and HTCs will recognize that their organization sucks at software, because it's not a software organization at its core. (Nokia spent tens of billions of dollars on Symbian and MeeGo until they finally got this lesson, and they ended up selling the phones to a software company.)
That's just the bargain Google struck with carriers and OEMs. They have the reins on Google services, but allowing OS customization was a concession to get third parties to adopt Android as the de facto iOS competitor. And OEMs and carriers love to push their own customizations to build revenue streams.
Funny that you mention that, since my Nexus 4 was upgraded to KitKat as soon as it was released, my first generation Galaxy S is on Android 4.3 (thanks to Cyanogenmod), while my home desktop is still on Windows XP (OEM), because Windows upgrades have been very expensive. The situation is much better with Windows 8, but coupled with taxes, I still would have to pay about 150-200 EUR in my country.
Microsoft's entire business model has always been OEM prebundling, in which updates happened by buying a new computer. That is why many companies are STILL on Windows XP.
Anything else is bullshit coming from young people that haven't felt how the industry evolved during the nineties. And Android is very unlike Windows because Android is open source, Android can be ported by third-parties and Android can be forked.
Then again, Android comes in different flavors while Windows does not. There is a trade-off there, and you'll usually get the same or better upgrade frequency from an Android OEM than from Microsoft.
For those interested in having the latest version (most consumers aren't even aware their device is outdated, or don't care) there's CM and similar options. Driver issues exist with some hardware (Samsung Exynos for instance, prime example), and that's a shame, but it's not much different from going to GNU/Linux on x86, which I find is comparable (more up to date, generally better performance and UX, non-proprietary, but some hardware incompatibility)
My peeve is that even when Windows came with all the bloatware crap that was bundled by OEMs there was a way to uninstall it. On Android you have all these apps (Facebook and the like) that you often can't uninstall without rooting the device. I guess that's just some of the poison that comes with it being open: carriers can do even worse than with Windows.
Microsoft used to strictly control the look and feel of the desktop, so OEMs were not permitted to add a new desktop or uninstallable software. The equivalent of TouchWiz wouldn't have been possible on Windows.
> Microsoft used to strictly control the look and feel of the desktop, so OEMs were not permitted to add a new desktop or uninstallable software.
That ended after the antitrust case. As part of the settlement, Microsoft was required to allow OEMs to preload anything they wanted on PCs. It doesn't matter if it's bloatware or a new shell -- Microsoft can't say no to any of it.
For example, Lenovo is now preloading Pokki on many of its Windows 8 PCs.
While the antitrust settlement was binding only on Microsoft, other OS vendors will now look at the Microsoft example to see what restrictions they're allowed to get away with.
The solution of course is for Google to charge a small fee, say $50k/yr to each OEM who wants to use Android on their devices (if the OEM can't handle that fee, then they have no hope of handling the huge manufacturing, support, carrier fees and other costs involved with launching a product).
This fee will guarantee them a warm body with a heartbeat on the other end of a dedicated number who will be their "relationship manager" who can run things down internally and manage bringing new OEMs on.
In reality, the RM would probably service 4-5 "accounts" at a time and these 4 accounts or OEMs would completely cover the cost of this employee or
Of course having 20 or 30 RMs banging against the Android team to get info and support OEMs doesn't scale, so for every 3 or 4 RMs you hire additional staff on the Android dev side to handle that relationship. And that does scale.
It's not a money maker for Google, but it covers support costs without turning support to OEMs into a cost center.
This would certainly be a solution for OEMs in this situation, but I wonder how much of G would even consider this a problem? Another way to ask this: how long will G neglect the needs of this market before we conclude that G doesn't consider this a market worth noticing?
It seems possible that a third party could provide many of the services OEMs require, but I suspect that organization would need to be either very skilled or somewhat well-connected to the relevant dev groups at G, or else it would find itself in this same exact situation.
How did Microsoft manage a zillion OEMs who had to have audited numbers and get invoiced, and Google can't manage a simple compatibility test and license agreement?
Never mind third tier phone OEMs, this is really hurting Google in tablets, where you don't need carrier retail channels.
Microsoft - Bill Gates - from day one believed they were in the business of selling a product to a customer, providing support to that customer, and they wanted that customer to stay with them forever. For the first 15+ years of Microsoft's existence, they had no monopoly position, and had to always be hustling for customers.
Google being a search engine, from day one never had to nor intended to provide support to any users. By dominating search usage (they owned search by what, five years in?), it guaranteed they'd own the search ad business by default (and of course they hooked up an excellent approach courtesy of goto.com), with that combination in place there would never be market incentive to provide good customer service to their ad customers and partners (and once that behavior was encoded into the company's DNA, they approached everything a similar way).
Sometimes Google seems more like a hobby farm (that happens to have a golden goose) than a properly run business.
Let me re-write that a little to make more sense "From Day 1 Microsoft believed their success was based on partnering/OEM with hardware and software companies to get their products to market. Based on that they built out the infrastructure to manage all aspects of partnerships. Google, however, started with a direct to consumer offering that had no need for a costly partnership management infrastructure. So both built up organizations to support their initial customer base and did so very efficiently"
If only there were a term for being so impressed with yourselves that you felt you could just ignore the little people. Maybe "focused on Q1 revenue goals"? Seems long, though. I'd be happy if I could get it down to three syllables.
> If only there were a term for being so impressed with yourselves that you felt you could just ignore the little people.
It nothing to do with people being "little" in Google's eyes. It's about them being numerous. According to Wikipedia, Google has 46,421 employees. Here's the reported number of users of various Google products normalized to "per Googler":
Imagine you had your own B2B business. Just you, no other employees. You have 107 businesses using your product. How much time would you be able to give each of them?
Say you open up a little hand-crafted artisanal phone business. Just you. You've sold 19,387 phones in the past few years, and you sell 18 more every day. How much time would you have for support for each of those?
You make software. Just a little hobby on the side you do by yourself. You have more than 32,296 active users of it. You're still cranking away on it, adding new features. How much time do you also have to answer the phone when one of those people has a problem? Keep in mind, none of them have paid you anything.
Let's say Google decided to drop everything: no development, no new features, no new products, no vacation, no HR, no facilities, no security, no R&D, no training, no nothing. Instead, every single employee from Larry Page on down devotes their full time to tech support. Users are pretty sharp so they rarely have problems: they only need to call for help once per year. If all every Googler did was get on the phone and help people, they'd still only be able to give each person 4.82 minutes.
You say it like it has to be one or the other, but why can't it be both?
Google has chosen businesses where most of their users aren't, as you point out, their customers. They're the product.
What made them interesting in the beginning is that their behaviors were different than the economic incentives. They were engineering-focused, so they made a great search engine, and didn't worry about revenue maximization. They were, as far as I can tell, really serious about organizing the world's information, and about not being evil. That helped them crush their competitors.
But as far as I can tell, they've become just another large company. Many thousands of people, interacting mostly with fellow employees. So the people, always numerous, have also become little. Which I think is always the seed of a company's downfall. E.g., Microsoft, which has floundered for how long now? Cut off the roots and the whole plant eventually dies.
My experience is that where they directly make money, they're very responsive -- their adwords team has been very responsive to us, a small startup. And I bet they're very responsive with the big OEMs.
They don't have much financial reason to put energy into small Chinese OEMs. Google wants to control advertising, and as long as they have relationships with the folks selling the premium handsets in china, they have most of the advertising $$.
I'd agree. On the ad exchange side, their customer support is the best in the industry. But then a lot of that DNA comes from DoubleClick where they have always needed to provide sales and support services to a wide variety of customers and partners.
Ironically, this actually isn't hurting Google in tablets. Over the past year or so, they've pulled in all of the bigger names (B&N, Kobo, Nabi, etc.) except Amazon and many of the budget OEMs as well (Archos, Coby, Ematic, ...). Maybe they're doing more outreach because Google is more concerned about tablets. It almost makes me wonder if the best way to get Google to pay attention to your phones is to make a tablet or two as well.
That was unexpected. Why would Google be so ignorant of Android China? China seems like such an obvious part of the mobile big picture that I can only assume "big" things.
In no particular order:
• Google's decision to leave China [1]
• Political interference to protect a Chinese mobile OS [2]
• Some sort of license restriction that doesn't effect Google services over the web but kicks in for mobile (completely a wild guess here)
• Political interference to keep Samsung out (so implausible I don't even believe it, but for the sake of discussion...) [3]
* A desire to protect Google's relationship with the big-name phone manufacturers: maybe the big players have hinted that Google getting too helpful to the small Chinese firms would tend to cool them on Android
* Concern that the Chinese manufacturers might accept Google's technical help, then refuse to implement all of the G-suite or otherwise violate the terms of the certification. With Google's weakened position in China it might not be able to get effective redress
* A specific form of the above: concern that Chinese-brand phones will contain trapdoors to allow clandestine Chinese government access to your Gmail etc. Google can't prevent this from happening on Chinese phones with "unofficial" G-suite access, but at least it can disclaim responsibility, never having certified the device
Wasn't the problem has to do with small vendors, not big names?
I would assume that was the issue? And if so, my guess is people just thought Chinese market was too complicated to handle (speaking from a Chinese perspective...) and particularly it was hard to deal with all the malware, links rewrite and law conflicts at all once, even with the powerful cash Google has. And since the Google China incident, Google doesn't seem to have any interest in the Chinese market anymore. It is a big market, but the effort to put into it to compete with local competitor is just way too much; plus Apple is already part of competition. I am actually amazed how Apple is able to succeed. To many people in China, owning an expensive phone like iPhone is a necessity (well, maybe I will call it a norm).
Right now I think mobile market is shifting interest to developing regions like South America and Southeast Asia.
Apart from Matt Cutts responding personally on HN, people trying desperately to sell you Adwords, and press releases; they may as well be a black hole.
I had a recruiter from Google call me, chat for 20 minutes, talk about my resume, then ask for my resume. I told her to bing my name and download the latest copy.
Same here. I told the person to type my name in google, with the word "resume" next to it, and that the first result would be my resume. She was astonished and asked how I could do this. I replied that I followed Google's own s/developer/webmaster/ guidelines. (Try it: google "chris mahan resume", it's still there, three years later, first result.)
I subsequently failed the basic algo questions, though, not having a computer science degree (thankfully--both for not working at Google and not having a computer science degree).
This was my reaction to the article as well. I'm still baffled why they don't take some of those billions and invest in a top notch customer service team. I'm pretty sure it would increase their bottom line about ten fold in less than three years.
Google loves things that scale, and customer support doesn't scale that well I guess.
Alternatively, couldn't they start a pay-per-minute service? I recently had an issue with GMail where I was receiving someone else's mail. After about a week I had enough data to source his phone number and contact him to rectify it. But I never had the chance to report the issue. Wouldn't have minded paying a few dollars to have that one rectified at the source.
Once you've lost someone's good will, it's very difficult to get it back. Google terrible, impersonal customer service is almost calculated to piss people off. Either they've calculated that they'll still make money when everyone hates them, or their business strategy is quite short sighted.
Google has about $1,000,000 of revenue per employee [1].
Assuming their support employees worked a typical number of hours and billed every hour they worked to a customer, to keep the revenue per employee at the same level they'd need to charge $480 to $550 per hour for support. And fixing complex bugs in complex software can take a lot of hours.
How many hours of developer time would you want spent on your issue?
Wonder why Amazon doesn't seem to have those problems. To me, they have just as many, if not more products to support and yet they're still known for their exceptional customer service.
Different types of support and different designs of company.
Call centre workers can process refunds, send out replacement items and do similar routine things - and they're not that well paid. It's unlikely there will be other more profitable ways to deploy them in the company. Programmers who can make code changes to fix faults in production software cost a lot more. And there will be other departments (which make a bigger contribution to the bottom line) trying to attract your most talented people.
By design of company I mean some tech companies only employ a small number of non-programmers, so they can afford to give programmer-level salaries, benefits and working enviroment to all employees. Google could give free food / laundry / busses / nerf guns to all employees and might think the equality is culturally important. Amazon on the other hand has radically different pay/benefits/working enviroments for programmers vs call centre workers vs warehouse workers.
AWS support is pretty terrible, my last ticket didn't get a response from them for 8 days. Though when I complained about it on Twitter they responded a few hours later apologising and offering a refund.
Slighlty OT: Anybody knows why installing an OS on mobile hardware cannot be just like installing Windows in the old days? First installing the base system and then drivers for 'non-standard' hardware components. Why do need hardware manufacturers and OS vendors to talk?
Getting the G-Suite seems to be a problem now but there are already some alternatives for each of the G-Suite's apps and I'd welcome any newcomer to this stage-so I wouldn't consider the G-Suite as the main problem with Android.
It is possible. For example Ubuntu is installable on some phones.
It is like Linux 15 years ago, there are a few good drivers and one has to check compatibility charts.
Other problem is hardware detection. OS on PC can detect what hardware is installed. On mobile you need hardware map. There is no way to have universal installer for all mobile phones.
There is no plug-and-play, you need to know if the hardware is there or not. If you poke something and you are on the wrong board, if could free/reboot/brick/whatever.
It is an ARM thing. The good news is, that there is a map and a single kernel can run on multiple boards. It is a recent development, before that, you needed a separately built kernel.
You can download source code and blobs for some Sony phones. Something things work, some require closed source blobs. Some things don't work because the driver providers don't allow Sony to redistribute them separately.
Android itself can be installed without the Google suite otherwise known as GApps. Actually most roms don't include GApps for this reason. Look at any AOSP based rom. Its just they want access to the Google suite and this requires OHA http://www.openhandsetalliance.com/ requirements.
The Android devices I've had the displeasure of using (Galaxy S2, Nexus 4) reboot/crash/hang on a more regular basis than any Windows computer I've used in the last ~10 years. I used WebOS devices for a while before switching to Android, and I don't remember the last time either of those crashed or needed to be power-cycled. My iPhone 1 was an unreliable piece of garbage, though, so I guess I can't specifically shit on Android here.
Anyway though, 'people don't want their phone to BSoD' seems a rather unsupportable statement: modern 'big seller' Android phones suffer from surprisingly poor quality control, especially if you install the OS updates on a regular basis. If they're not introducing some sort of hang/crash bug, they're breaking one or two of the built in applications. I remember one of the recent Nexus 4 OTA updates broke one of the built-in preferences panels under Settings - how do you ship an OS release without testing that?
Ultimately cool new features and apps are more compelling to people than actual stability, judging by how many of my smartphone-owner friends complain about phone crashes and hangs. They're still using those phones and buying the latest models...
Same here. Two different top-of-the-line Verizon/Motorola phones, both have an average uptime of a few days. Generally due to a crash, but sometimes because performance is so terrible I have to reboot it. And thanks to the carrier's lockdown (for stability, they'll surely claim), I can't actually debug any of the performance issues.
If anybody knows of an Android phone that actually works reliably out of the box, I'd be excited to hear about it.
The Google apps are significantly better at present. What is needed is a community effort to work on one set of replacements that work well, rather than the fragmented alternatives.
Indeed and this set could be consolidated around an ownCloud API. I would even go further make them isolated from google's API by default with options to enable tight integration with the platform, if you wish. That is for instance, an address book that does not share your contacts and a calendar app that might share it's event if you explicitly say so, an A-GPS that doesn't report automatically your position to Google ...
Let's prioritize... Help out a large OEM with some level of gravitas and reliability or 10's to 100's of bargain-bin-targeting mom & pop shops that are unwilling to put their balls on the line to get up to speed on making globally-commercial products. Then you have the morass of cheap China tainting your platform. Google loves Samsung because they understand, slightly, how to balance silly, shiny shit with passable hardware. Imagine the backlash when millions of consumers think "terrible piece of crap" when someone says, "Google-compliant, Android-based device."
AND THEN those same OEM's will come back and tar Google (again) for their failures. I wouldn't rush out to prop up everybody with a cousin that knows a bad batch parts fence in Shenzen. Let the creme rise to the top first.
So I wrote the original post, and I'm always a bit afraid to wade into the comments, but I wanted to respond to this one.
Your point is that Google is prioritizing correctly. That they are just going after the big OEMs first and then the guys I'm talking about are too small to bother with. About right?
I didn't hammer the point in this post, because I've made it a few other places, but all of these OEMs collectively are bigger than any other single OEM. Collectively, this group produces between one half and two thirds of the world's phones. Most are very small, but many are now huge. Google seems to be focussing on just one or two. I think with a small amount of effort they could tap into the whole audience. It wouldn't require building a huge support structure. And the numbers are huge.
Thanks for the wading and additional, nuance-providing details. It would still seem that they would first focus on a single tier of support focusing around their favorite few largest OEMs to set the guide for the others. I'm not saying top 20% of devices, but something along the lines of focusing on the an 80-20 principal. Then, if you see a star being born that will disrupt your top picks, bring them in. Market dominance can be sustained that way. Those OEMs aren't going to be making iOS or Lumias. "It's not nice, it's Google." (TM)
Can anyone explain how I can just install the G-suite on Cyanogenmod then? Is it blessed by Google? Or does it recognise that my phone is appropriately blessed via the bootloader?
Read his question again. He didn't ask how to do it, he knows that. He asks why it works.
I was wondering the same thing when I read the article - my best bet is that the idea of ‘cracked’ G-Suite keys is a journalistic fiction invented for the sake of the story.
Not that it's legal for them to distribute it when they are not certified, but I don't think that there is actually a key required.
You can open the .zip and replace the keys or clockworkmod which cyanogenmod uses allows it to copy from recovery by ignoring the fact keys are different from platform. I guess since they just provide binaries and not source, and google makes money when you use their apps and provide them data they don't care that other projects have GAPPs for dl
The idea is that you can backup the Google binary blobs from your stock OS, install CyanogenMod, and then restore your own binary blobs. In practice most people just download GApps from illegal sites because Titanium Backup has such an awful UI. But if Google ever went after the download sites you'd still be able to use the backup method. Which is probably why Google doesn't bother with the download sites.
> The site(s) you're downloading it from are committing copyright violations and are subject to a DMCA takedown.
Yes, that is technically true. But in practice, it's not like Google just haven't noticed yet and are likely to take them down any day now: the current situation (cyanogenmod ships without gapps, but google won't prevent goo.im hosting them or cyanogenmod linking to goo.im) was worked out in 2009 after that C&D, and Google seems to have been fine with it for the last 4 years - they even let goo.im have an app in the play store - even if, officially, they still reserve their right to assert their copyright in the future.
(The wording of the license for gapps now specifies that the licence is scoped to the device[1], so an end-user who installs gapps (e.g. from goo.im) on cyanogenmod is not committing copyright infringement if their phone came with gapps (i.e. not kindles etc.))
Yet Google have permitted goo.im to continue making the apps available for download for years now. The site is registered via a US domain registrar to an address located in the US, so the individuals in question are not beyond the reach of a DMCA request.
Google obviously decided some time ago that cutting off CyanogenMod from the Google universe would not be in their best interests, whatever the strict legality of the situation.
this
"Google offers the Android codebase for free download, but this is only part of the software needed to make a phone viable"
is simply false. You have everything needed for a "phone" and even for a "smartphone", you simply miss google services that might (West in particular) or might not (East in particular) be perceived as a must have by consumer.
See for example all the android-based phone that ship with alternative markets, maps and similar services.
Nobody, not even Oppo with their CyanogenMod-based Android, is selling an Android product in developed-world markets without Google Play and the rest of Google's app suite.
The real alternative is expensive: Build your own ecosystem, as Amazon has.
You seem to have come up with the refutation to your claim, but then ignored it anyway. The Kindle Fire [1] does run Android and does not have the Play Store - it has the Amazon app store instead.
Sure - the 'smaller-brand' ICS tablet I have got its apps from the Amazon App Store almost exclusively, with a few sideloads for good measure. Figured out how to get Play services on to it, and have the best _and_ worst of two worlds :)
("Figured out how to" is its own death-knell, I know, in some circles. Not that clueless.)
I feel like this author just doesn't know what he's talking about and is just ranting on google because AOSP != google experience. He makes no mention of Open Handset Alliance http://www.openhandsetalliance.com/. This would be the first place to start.
I'm the author. And I do know what I'm talking about. OHA doesn't even begin to solve the problem at hand. The subjects of my piece have access to AOSP, and could join OHA if they wanted to, but the fact is that those are not enough to support their products. These OEMs sell into incredibly competitive environments, and many of their customers want the full Google Experience and G-Suite. They'd be willing to pay something for it. My argument is just that it is in Google's best interests to become more engaged with this community. Nothing to do with how I feel about AOSP.
Author didn't say specifically what kind of requests Google didn't respond to. It could easily be the case that given EOM's device isn't good enough to pass Compatibility Test, and instead of fixing the issue EOM would contact Google saying "our phone doesnt really need this feature that your test requires, but you still give us keys to GSuite." Of course Google won't respond to such inquiry.
Why not offer a link to the Gapps download with instructions for installing them like CyanogenMod does? It wouldn't be that hard to streamline this for novices on first booting their phone and therefore bypass Google's bureaucracy.
When this happened to Microsoft, certain strategy-oriented parts of the company were happy to see Windows spread everywhere as quickly as possible and deal with piracy later, whereas others were more concerned with the bottom line and how to extract revenue immediately. Google will have to deal with the same political questions.
It's funny how Android even has its own versions of OS/2 in the form of BlackBerry 10 and Jolla's Sailfish -- that is, competing operating systems that proclaim to be technically more advanced while offering reasonable app-level compatibility with the dominant system. (I suspect that BB10 and Sailfish will share OS/2's fate in a few years. Too little user benefit against the high cost of lacking access to the dominant vendor's services.)