Next week, Google is holding an event in San Francisco to announce (we’re told) new Nexus phones along with a finalized Marshmallow version of Android along with some other hardware. The Nexus phone line remains one of the oddities of the global smartphone market and there have been rumors for some time that the whole program would be discontinued. And yet, here we are, about to see not one but two new Nexus phones announced. Why?
Nexus doesn’t seem to be working
We’re already a long way from Google’s original vision of Nexus devices. Back in 2010, Google began the program with the launch of the Nexus One, which was intended to bypass traditional carrier channels by selling exclusively online. However, the phone was more or less a flop, and Google itself said, when it discontinued online sales a few months later, “While the global adoption of the Android platform has exceeded our expectations, the web store has not. It’s remained a niche channel for early adopters, but it’s clear that many customers like a hands-on experience before buying a phone.” That remains a very good summary of the fate of the Nexus program since then, too. Google has continued to struggle to get wireless carriers to support and stock its Nexus phones at retail, especially in the US and, as a result, buyers have been limited to two fairly narrow groups.
Those two groups have been developers and a niche segment of non-developer users who care enough about using stock Android devices they’re willing to pay full price for them, since Google has never been able to take advantage of carrier subsidies or the more recent equipment installment plans. It gives away a fairly large number of Nexus phones to developers at I/O and through other programs, while selling them at close to cost through the Play Store to both developers and others who are interested. These two groups, even combined, are small. Small enough that sales don’t really register on most analysts’ smartphone trackers.
For a while, Google experimented with selling stock versions of non-Nexus handsets as a way to help developers test new versions of Android (and their apps) on top-of-the-line Android hardware – the HTC One and Samsung Galaxy S 4 and several others. In many ways, this seemed like a better solution to the problem the Nexus devices were trying to solve, but Google abandoned the project, likely because OEMs were unwilling to participate in this way. The big advantages of this approach, though, were that Google got to use high-end devices (whereas Nexus phones have often been mid-tier devices) and leverage the scale vendors brought, without showing favoritism. By contrast, the Nexus program has historically favored a single vendor each year, in the process alienating others.
Interestingly, the core value proposition of a stock Android experience has been taken beyond the Nexus experiment during Google’s ownership of Motorola, which has largely focused on fairly stripped-down flavors of Android in its smartphones over the last several years. However, given that company’s flagging fortunes, it’s clear this model doesn’t have enormous appeal even when it’s sold through traditional channels. In a world where almost any Android phone can be made to run stock Android if you want it to, especially if you’re one of the users most likely to want that experience, Nexus devices are increasingly serving just the one constituency: developers.
A new role for Nexus – and new model?
Two interesting things have changed in the past year that may affect how Google views the Nexus series: the launch of Google Fi and the rapid adoption of installment plans and device leasing programs by US consumers. Google Fi, because of its unique technical requirements, currently only runs on a single device — the Nexus 6. However, that phone is unusually large even by today’s standards, and a single device, especially one you have to pay for out of pocket, is a poor solution for a wireless service in today’s market. If Google Fi is to succeed, it needs to offer a wider range of phones and it needs to find ways for consumers to pay for them easily. Edit: Google does now offer financing for the Nexus 6 when purchased as part of Google Fi.
Enter the two new Nexus devices on the one hand and the potential for an installment plan model on the other. With two new Nexus phones (both of which will presumably be compatible with Google Fi), Google will have three devices it can offer that work on the service (assuming the Nexus 6 remains available, perhaps at a discount). Google now offers financing for Nexus phones through Google Fi, so assuming this model is extended to the new phones, that helps to solve that problem. The remaining major challenge, however, would be the retail problem it identified in that 2010 statement. Given Google’s lack of success getting carriers to stock its Nexus phones, how will it allow them to test these devices?
A showcase for new Android versions
The other role the Nexus devices continue to serve, of course, is the only way Google can truly get its new versions of Android into (at least some) consumers’ hands when it launches. Like Apple, Google now announces new versions of its mobile operating system in the summer, followed by availability in the fall. However, unlike Apple, which has seen over 50% uptake of its new version of iOS within the first few days just this past week, new versions of Android tend to find their way into consumers’ hands extremely slowly. The big bang announcement of a new flavor of Android in the summer is often followed up by a whimper in the fall, when the new version is announced but available to almost no-one. The Nexus phones continue to be the way developers get access to these new versions, but they’re also pretty much the only way ordinary users can get access in the first few months. This past year, Lollipop didn’t start showing up in Google’s own stats for adoption until several months after it became available.
A program that needs a new purpose
Many developer programs provide hardware for testing purposes but, in the vast majority of cases, this hardware is either the standard version ordinary customers will also end up buying, or custom, stripped-down hardware for the sole purpose of developer testing. The Nexus program is odd in that it combines this same core purpose of providing devices to developers with the fanfare and branding of a consumer product, albeit a niche one. At some point, Google needs to apply the lessons it apparently learned but didn’t heed after the Nexus One launch and find a new purpose for the Nexus program. Perhaps Google Fi will provide that purpose and perhaps a device financing plan can stimulate sales. But, so far, there’s little evidence this year’s Nexus phones will fare any better than previous year’s versions.
49 thoughts on “Why Does Google Still Make Nexus Phones?”
I think Google are using the Nexus program to
1- show the way to OEMs. A few years back it was “cheap”, last year it was “big”, this year it will be… feature rich ?
2- have a “pure” device that showcases both bare Android’s look&feel and capabilites, and advanced hardware support (Fi, EAP-SIM…)
3- have an iPhone-equivalent device which they directly control and quickly update.
The lack of an overwhelming need for the Nexus is a sign that OEMs are doing their job ? Paradoxically, I am personnally more interested than ever, because of security updates. It seems the price will be ridiculous though, I’ll probably just buy something sensible and pass it down next year like always.
I don’t think Google want to sell that many Nexuses (Nexi ?). It’s there for those who care about any of the above, especially since Google Play Edition phones have been discontinued and Android Silver never… materialized (ah !).
I’m not foreseeing a big push, especially since they still go to the trouble of working through 1st-tier vendors instead of going directly to Pegatron et al, which they certainly could do. That’s a big difference compared to Microsoft with Lumia/Surface.
BTW, the “unusually large” Nexus 6 is 1mm longer and 5mm wider than the iPhone 6+. Fairly negligible, especially for half an inch more screen diagonal. For reference the iP6+ is 20mm/11mm larger than the iP6 for 0.8″ extra screen.
See, there’s a large part of the problem, right there. Each time the Nexus program fails to find success, Google tries to re-jigger it, leading to confusion. No one knows why they should bother to look for a Nexus phone! (See the OP)
Saying that Google wants to try different things out with the program and/or give examples to the OEMs is simply justification after the fact. “Their plan didn’t work.” “Well, what they were *really* trying to do is…”
One of my friends who’s a Google engineer got a Nexus One when it first came out and he told me “This is it! The iPhone is done!” (“…Now witness the firepower of this fully armed and operational corporate juggernaut!”) LOL, nope! Time for another plan!
As for the failure of the Nexus program being a sign that Android OEMs are doing their jobs: That must be why except for Samsung they’re all losing money, and even Samsung is in trouble with multiple quarters of declining profits and market share?
Now, it is a shame that Nexus devices aren’t more popular because the OEMs are so horrible at software updates, leading to the majority of Android devices having unpatched security holes. Since the Nexus program never set the Android world on fire I keep hoping that phones shipping with Cyanogen Mod will get more popular instead, since they similarly decouple the OS updates (with attendant security patches) from the OEMs and carriers.
P.S. Many people do think the iPhone 6+/6s+ *is* too large. That would be why the more normal-sized 6/6s were introduced and made available at the same time. So yes, the Nexus 6 is indeed “unusually large” as the *only* available offering with the rest of its particular feature set.
I don’t know where the “all Android OEMs are losing money” myth comes from.
1- Some figures here: https://theoverspill.wordpress.com/2015/02/09/android-oem-profitability-and-the-most-surprising-number-from-q4s-smartphone-market/ . Granted, not a lot of money, but mostly they do make some.
2- They’re not in the business for fun, so either they are making money, or they think there’s a good chance they will, soon. Except HTC who has no other business to be in.
3- The other, unlisted OEMS (Xiaomi, the hordes or 2nd/3rd tiers local ones…) are probably making more money, proportionally, than the big ones.
Re the Nexus, individual employees are usually not that aware of what the big picture is. I remember a microsoftie friend of mine parading about the spec sheet of a tablet from some obscure OEM 2yrs after the iPad’s launch, also convinced this was IT for that. I’m really unconvinced the Nexus marching orders are to sell lots. Why did Google sell Moto if they want into hardware ?
From what I hear about these years Nexus phones & Androids ongoing fragmentation, security and update problems, this is likely going to push more people to. IPhones. The sad part is Android has a lot of great features that every iPhone user (and Apple Geniuses) Ive shown thinks ate super cool. But the weaknesses of Android, the plethora of devices, and poor marketing sabotages them
Fragmentation is a red herring: it’s a synonym for diversity, hence not an evil per se. Nobody ever bemoaned it for DOS nor Windows…
But Windows handles that diversity in a much more elegant way, with an HAL and drivers where Linux/ARM is just one big messy kernel. I’m thinking Google need to do what MS did a while back: hit the reset button on the architecture and modularize it with an eye to security (=updates), which probably means isolating the phone part too, so updates don’t have to be carrier-validated.
They’ve been doing it, but slowly and partially. Having that kind of issues when one the major characteristics of a modern/mobile OS is that apps run in one big sandbox and only access relatively few APIs is… negligent.
Considering the way OEMs garbage up android the google nexus is the only android phone i will buy. Until now:
Google needs to continue the Nexus program but needs to support phone hardware longer. There’s no reason for the 2 year(or 18 month) cutoff on hardware support anymore.
Which 2 yr cutoff ?
quick google search
not hard to find. Google typically only gives you 18 months of support per device. Apple is now up to 4 years.
Not hard to compute either: 2015 – 2012, that’s 3 years, not 18 months, not 2 yrs.
Also, that’s for the bare version of the OS, not for Play Services, nor the PlayStore-based tools and apps, which keep on being updated. That’s an important distinction: Android didn’t need an OS update for Wear, Pay, Wallet… as opposed to iOS, because those are Play Services-based.
I know it’s a bit hard to understand: while iOS takes away features on older hardware, Android actually adds features to older OS versions. Makes most of the talk of “no updates” nonsensical, which might get through, some day…
android most assuredly requires updates for the base os in older version. just look at the security holes that exists in hundreds of miilions of phones now that are outside of the apps you refer to. The base os is getting patched for 4 year old gear in apple land..android……not happening. You can patch the apps all you want but if the foundation if insecure who cares about the apps.
The only vulnerability I’m aware of the targets the core OS is the stagefright one. All the other ones target stuff that’s at the Play Services or PlayStore level, ie can be/have been fixed independently of OS updates: browser, webview, lockscreen…
The stagefright one *is* bad, it requires either completing switching off MMSes if you don’t use them (that’s the overwhelming majority of cases) or at least disabling preloading if you do use them.
And there’s this rather massive flaw affecting every unpatched Android version:
The text one:
1- I mentioned it,
2- it’s not text it’s MMS
3- what they say about it: “If this sounds familiar, that’s because this Android flaw is somewhat like the recent Apple text hack.”
The lockscreen one: fix being deployed, change lockscreen if fearful.
It’s not as if Android were the one recently infected by hundreds of xcodeghost apps :-p
A bit off-topic anyway, this is about updates not security (linked topics, but not the same).
Somewhat like, but with major differences.
Xcodeghost is an example of why Apple’s approach is more secure. The isolated xcodeghost incident is the everyday norm in Andoidland. A hacked compiler downloaded from an untrustworthy source is the least of most iOS users worries.
Even antivirus software on Android is suspect. I would have thought an Android apologist like yourself would be aware of that.
” while iOS takes away features on older hardware..”
Only one ? Force touch, live pics, touch ID, …
Do those count? They all require hardware that older devices don’t have. I think the question was about software features. I’m actually pleasantly surprised that we can run iOS 9 on our iPad 2s, from 2011. I think even you might be forced to admit that’s pretty good support for older devices. Yes, I’m sure some things will be slower here and there, but as we’ve discussed before, I’m not terribly concerned if it takes 0.3 seconds more for the camera app to open, etc.
Why not ? You have the same issues with apple wear, apple pay (requires 9/2012 iP5, equivalent requires Android 4.4 ie 5/2012 Galaxy S2…)
I could go on with the huge list of minor stuff that only works on more recent devices. Updates maybe, but partial updates.
I’d say both ecosystems are at par.
There is a huge difference between features that are ‘taken away’ and features that with specific requirements requiring new hardware.
Your arguments are utterly dishonest.
What features does Live Photo require ? A camera ?
There’s no need to be disingenuous, you know better.
Sure, but none of that is iOS taking away features, there are new features that require some level of hardware integration which means those features can’t work on older devices. That’s quite different from what you originally said. Both ecosystems are good if they deliver what you need. Android can’t do that for me, but if it does for you, that’s great.
All features require some level of hardware integration. Apple chooses to skimp on RAM, NFC… to maintain its users on the upgrade treadmill.
That explains why, but doesn’t explain away, why “iOS9” is degraded and loses features on older hardware, while older versions of Android usually support newer features via Play Services or the Appstore.
OS versions don’t mean the same thing on iOS and Android, and iOS has under-the-radar fragmentation though missing hardware and software features on older models.
Now you’re having a different discussion, with yourself apparently. As an Apple user I don’t feel like I’m on any kind of upgrade treadmill. Quite the opposite. My iPad 2 purchased in 2011 gets iOS 9. Now obviously I don’t get every new feature because my device doesn’t have the necessary hardware. Live Photos for example, my understanding is there’s hardware in the new camera that has something to do with it. Of course back in 2011 there was an iOS app that did ‘Harry Potter’ photos, but it doesn’t seem to be available today, maybe it wasn’t all that popular, or maybe doing it all in software had some issues, or maybe they didn’t have a business model.
Back to iOS 9 on my iPad 2, I’m gaining lots of features and not losing any. What is happening is some new features can’t be added to my device because of hardware. You’re revealing a very strong bias when you choose to say iOS 9 is degraded and loses features. That isn’t reality.
Android’s model of more separation between software and hardware, Play Services and OS, certainly has some advantages. One system isn’t better than the other, they’re offering different experiences.
Saying OS9 loses features on older hardware is showing bias ? Let’s ask experts:
“iOS9 on the iPad 2: Not worse than iOS 8, but missing many features”
Split View, Slide Over, and Picture-in-Picture multitasking.
The new Spotlight screen, predictive Sir, and third-party Spotlight search.
Public transit directions for Maps.
Handoff features in Continuity
Support for OpenGL ES 3.0, the Metal graphics API, 64-bit ARMv8 apps, and TouchID/Apple Pay.
And you still want to pretend you’re getting the real iOS9 ?
Your argument proves that the ONLY difference in hardware between a three-year-old Android device and a brand new Android device is that the brand new device merely SUCKS FASTER.
‘Innovation’ is a word Android users don’t understand.
Here’s a clue: repeating the same derp over and over doesn’t make it true or convincing.
‘Innovation’ is a word Android users don’t understand.
That’s it in a nutshell. Mr. O can’t comprehend why an OS upgrade doesn’t magically install a 64bit processor into an iPad 2. He doesn’t get why Apple would require the encryption and security features of their 64bit processors to implement features like Metal, TouchID, and HealthKit.
He fails to grasp that Apple chooses to implement features so that they perform acceptably, rather than follow Samsung’s lead and shove Multi-Window Multitasking onto a device that is so slow and crappy that the feature provides no effective advantage over using a single app at a time.
He believes that his Android devices are already good enough, and that Apple is evil for producing new devices that differ in capability from previous devices.
Just wait until Google tries to copy some of Apple’s recent innovation and Google finds that 32bit support won’t cut it. Suddenly Google will be evil for not completely supporting Mr. O’s legacy device.
Actually, split view would mostly require RAM, but that would be $9 less profit/device for Apple. Same for Metal, there could be a 32-bits version of it, but that would take away one reason to stay on the upgrade treadmill.
Apple don’t look at acceptable performance, updates are mostly PR: “iOS 7 users should spend some time doing a deep think before clicking “Download and Install.” http://gizmodo.com/should-you-update-to-ios-9-on-your-iphone-4s-1731478418 . Again, it’s not me saying it.
Have you ever used a Galaxy Note ? Guess what, it came with enough RAM for multitasking… 3 years ago.
“Actually, split view would mostly require RAM…”
I bow before your hardware and software engineering skills. And your Android fan fiction is second to none.
“Apple don’t look at acceptable performance, updates are mostly PR…”
How can I help but be in awe now that I know my iOS upgrades are mostly PR? Of course, the gizmodo article explaining why some people still using 4yo phones might not want to upgrade is an earthshaking revelation. Utterly earthshaking.
obathelemy, your remarkable demonstrations of the superiority of all things Android leaves me bewildered. I’m off to the real world where everything makes sense again.
What do you think, that the inactive window in a split screen actually needs CPU cycles to… be inactive ?
Well, apparently the Apple fiction is that iPhone 4S run iOS 9. Apparently, they can in you like lag and missing features, but really shouldn’t… Who’s in the fiction world, again ?
As opposed to an extra clean 32-bit implementation which would be craaazy ? :-p Indeed, trying to support older iOS devices would be… what would it be again ?
I’m not seeing where innovation comes in ? Low-level graphics API that we’ve had since DOS days ? Their resurgence initiated by AMD’s something something ? Oh, their use in Mobile. Let’s call that a revolution: applying desktop techniques to Mobile….
“I’m not seeing where innovation comes in…
Of course you don’t. The underlying issues of doing it right are just PR to you. Security and performance are just PR to you.
The same way you declared below:
The only vulnerability I’m aware of the targets the core OS is the stagefright one… The stagefright one *is* bad, it requires either completing switching off MMSes if you don’t use them (that’s the overwhelming majority of cases) or at least disabling preloading if you do use them.
And that is just plain wrong. But because it involves your beloved Android, lets just deflect. Its only bad PR.
For your info:
“The scary [MMS] exploit, which requires knowing only the victim’s phone number, was developed by Joshua Drake, vice president of platform research and exploitation at mobile security firm Zimperium…
…There are many potential attack vectors because whenever the Android OS receives media content from any source it will run it through this framework, Drake said. The library is used not just for media playback, but also to automatically generate thumbnails or to extract metadata from video and audio files such as length, height, width, frame rate, channels and other similar information.
This means that users don’t necessarily have to execute malicious multimedia files for the vulnerabilities found by Drake to be exploited. The mere copying of such files on the file system is enough.”
So basically, a four-year-old iPad running iOS 9 and a slew of xcodeghost compromised apps is still more secure than ANY Android device that isn’t running the most recent fully-patched Android. We both know that is MOST everyone in Androidland.
So, why rub this in your face? Because you seem to believe that implementing multi-window multitasking securely on a mobile device is a simple matter of allocating CPU cycles. There are right ways to do things, and then there’s the Android way. Who knows what kind of attack vectors were exposed by Samsung’s go-it-alone approach to multitasking.
Pretending? I never said anything of the kind. We’ve established that older devices can’t run some new features due to hardware issues. But that’s not the same thing as iOS 9 taking away features. Nothing was taken away. Much was added in fact, but there are some new features in iOS 9 that I won’t get on my iPad 2, that is true. Now here’s the upside, my iPad 2 from 2011 is running iOS 9.
Interesting that you didn’t mention this part from the same article about iOS 9 on the iPad 2: “for all the stuff you’re missing you’re still getting a whole bunch of the new features we highlighted in our iOS 9 review.”
Be happy then, you’re getting a whole bunch of stuff, ie almost a real iOS9 :-p
I am happy, thanks. You are clearly not.
Indeed. I’m so sad with my 3yo tablet, its split screen multitasking, pen… :-p
And he has the security enhancements that no four-year-old Android will EVER receive.
And he’ll need them ! https://www.google.fr/search?client=opera&q=apple+hack&sourceid=opera&ie=UTF-8&oe=UTF-8#q=apple+malware+outbreak
Sure hope your Samsung tablet is running at least KitKat 4.4:
We could do this all day, but the difference would be that an incident like xcodeghost has opened no serious attack vector AND has nothing to do with any underlying iOS deficiency requiring a patch.
The steaming pile that is Android on the other hand…
iOS 8 & iOS 9 both added features to my 2013 iPad Air (which came with iOS 7). Nothing was taken away.
Nice Android fan fiction.
Are you getting all features that iOS9 brings to the Air 2 ?
They are different hardware. Why should I expect all the same features?
Your claim is that I LOST features I once had. That is utterly false. Rather, I have experienced performance improvements, better battery life, and many new capabilities that weren’t available on the device when I first bought it.
You love your Android. We get it. But to claim that Google advances older hardware while Apple screws theirs is ridiculous. Quit while you’re behind.
I’ve always just sort of figured t was their most direct form of market research or focus group for features–which are used and how much.
I have bought 3 Nexus phones in the past for $50 or less with carrier subsidies. The Nexus 6 is on Verizon’s website today at $27/month as part of their installment plan. So, I guess what I’m basically saying is your research and fact checking sucks and that really undercuts the confidence anyone should have in the conclusions you draw from it.
Haven’t many phone makers (Samsung excepted) moved toward the “pure Android” Nexus’ experience rather than away from it? Perhaps the phones themselves aren’t big sellers, but I’d say they have influenced the market, positively.
It’s not so much about being “pure”, as about updates.
Staying pure is making a lot of sense these days, custom launchers, tools and apps don’t really bring much to the table (unless you got specific hardware features to support: pen, touch id, force touch, …). Some OEMs are even letting us uninstall crapware.
Plus the Android world is wonderful: you can get your virginity back, just switch out the OEM stuff and put AOSP’s or Google’s back in their place (lockscreen, launcher, notifications, keyboard…). They’ll still be running on top of a customized version of the OS, but you will only see the difference in very small details.
On the other hand, the unwashed masses (and many of the supposedly washed ones) really don’t get that Android is layered so OS updates/versions are way less meaningul than on iOS; and core OS vulnerabilites are an issue.
When the iPhones with big screens came out last year, I returned to the Apple fold. But for two years before that, I was extremely happy with my Nexus 4 phone and liked it more than any Android phone under $400 on the market, and more than the iPhone 4.
I never get why people are dismissive of devices based on their sales. The simplicity of the Nexus 4, the automatic updates, the safety — it was a great package. But Google barely tried to market it. Another Google mystery play.
I hate when people comment on the “uptake” of new Android releases as opposed to iOS like its the users making the decision. When iOS has an update available it obnoxiously reminds you, even defaulting to the update screen when you go into the settings menu. I have an HTC one and I got lollipop a mere 6 months after its release. Some of that I could blame on HTC but I had friends with international and att Ones who still got the update before me and my Verizon phone. Why the nexus program? Well one reason is because some of us are tired of waiting for carriers and OEMs to “evaluate” everything (ie, add their BS to it) before its pushed to our phones. I JUST got my stage fright patch less than a month ago. A serious vulnerability and HTC still can’t push a timely update. Nexus is for people like me that are tired of being at the mercy of carriers and OEMs when its time to get the latest and greatest our OS has to offer. Now if you’ll excuse me I need to get my free wireless charger by signing up for Samsung Pay on my Veri…oh wait…its still being evaluated.
I got a Nexus in 2013 for several reasons. I wanted an unlocked phone so I had freedom to choose a less expensive carrier and a 32gb Nexus was also several hundred dollars less than comparable phones. It saved a LOT of money. Plus I didn’t want carrier bloat ware & I wanted to make sure I got the latest updates from Android. Part of Google problem is they announce a new OS & most users can’t get it for a year. Most people still don’t have Android 5.0 as 6.0 is being released. But given phones changing sizes and features, I’m no longer comfortable buying a phone online only without seeing & touching it. I want to see if that device is too big to fit in my pants or use comfortably. With subsidies ending, an affordable, high quality phone could succeed but Google seems to do very little to market them in terms of advertising or PR. Outside of tech blogs, you’ll likely hear nothing about these new devices or Android Marshmallow