We Can Now Estimate Global Android Forked Installed Base ie AOSP Devices vs 'full Google' Android

Google just announced 2.0B total Android users worldwide, at the end of last week.

 

This is a big milestone clearly. It also is not a smartphone Android count: it includes tablets too. But it differs considerably from my count of 2.6B total Android smartphones in use. I've said many times before that in my count, I include all Android manufacturers, both those who offer full Google Play and 'official' Android users, and those who provide the Android-compatible 'forked' usually cheaper phones that run on AOSP (Android Open Source Platform). I did this because up to today, it was impossible to separate out the two flavors of Android, so I just reported the top-line number. Most phones sold for example in China and India run AOSP while nearly all 'Android' smartphones in the rich world are full Google Android phones. Obviously in rough terms, about two thirds of all smartphones sold worldwide are 'official' Android and one third are 'forked' Android or AOSP versions.

We do not know the proportion by tablets, but it is very likely that the tablet market strongly mirrors the smarphone market. This means, that for the first time my consultancy can give a rough estimate of the total Android market separating out the AOSP devices. We also can then calculate out the smartphone market and its shares. Now that we have this data, I will be including it into the global market share installed base numbers as I publish those.

So here is my rough math. I estimate that out of the about 450 million total Android tablets sold so far (through Dec 2016) about 75% are still in use (tablets have a longer life span than smartphones, they are replaced less frequently than phones; and for comparisons, I estimate iPads have still 80% in use). So I get 340 million active Android tablets in use. When we add that to the Android smartphone count of 2.6 Billion, we arrive at 2.94B total Android devices (including AOSP) in use.

From the 2.94B total Android installed base, we see Google only 'sees' 2.0B ie 68%. That tells us that about 32% of all nominally 'Android' devices that have been sold, according to public sales data, are not visible to Google, so they are not 'official' Android. They are thus AOSP devices. I will use the assumption that this percentage will be the same for both smartphones and tablets (until we get any better data) and thus, we get a rough count for the Android 2.6 Billion installed base smartphone market: It splits 68% official Android (ie Play Store compatible) at 1.8B smartphones; vs 800 million AOSP devices whose owners would typically get their apps from a local branded app store or some national app store.

For context, lets remember that Apple iOS has about 600 million iPhones in use (and 300 million iPads plus the last 100M are miscellaneous other iOS gadgets like iPod Touch media players, Apple Watches etc). So out of the 'real addressable app store' ecosystem approach, Google-managed part of Android is currently pretty much three times as large as Apple's iOS in terms of smartphones. (that being the primary focus for my blog, ie it is the real 'mobile' opportunity we track). For the full-portfolio digital device scope, official Android is about 2x larger than iOS, when tablets and others are added. Note that as consumers, most who own an iPad will also own an iPhone, etc, so the unique smartphone market is also nearly the same as 'total unique users' on that given platform.

One last observation, the total Android 'potential' market beyond smartphones is 2.94 B and shortly 3.0B when AOSP devices (smartphones and tablets) are included. This is the potental that Google may be able to reach in future editions and on the migration and upgrade paths of those consumers. Apple's 'ceiling' is pretty close to the 600 million it now has, and is trying to sell SECOND gadgets to existing iFans, ie Apple Watches and new Macbooks etc, to existing iPhone owners.

I figured my readers were eagerly hoping to get these numbers; I certainly was thrilled to finally be able to estimate the AOSP share out of total Android smartphones.

(I will soon do the market share numbers for Q1 of 2017, but that race is no longer in any way dramatic...)

PS for those who need 'all' the data on handsets, remember the brand new TomiAhonen Phone Book 2016 (released every 2 years) came out just last December so the numbers are quite fresh covering the end of the year 2016. See more about TomiAhonen Phone Book here.

Comments

[Abdul Muis]

Tomi,

Google didn't say that Android reach 2 Billion user. Google say it **PASSED** 2 billion user. So, it perhaps 2.1 billion, or 2.2 billion, or maybe 2.6 billion.

[Tomi T Ahonen]

Hi Abdul

Good point but as the previous announcement was 400 million less, we can be pretty sure they 'just' passed it and have now about 2.01B or 2.02B users, not 2.1B or 2.2B haha..

Tomi Ahonen :-)

[chithanh]

Very nice analysis.

About the "official Android" vs. AOSP ratio in tablets we might not have to guess.

According to Strategy Analytics, Q4 2016 tablet market was 6.7% Amazon and 28.5% white-box manufacturers, probably a large part of them Chinese AOSP. So if Strategy Analytics does not severely miscount white-box tablet sales, that would put the upper limit of AOSP tablet share at around 35% and the real number would be somewhat lower.

https://www.strategyanalytics.com/strategy-analytics/news/strategy-analytics-press-releases/strategy-analytics-press-release/2017/02/02/tablet-market-falls-9-in-q4-2016-with-apple-samsung-down-double-digits

[Cesar]

Are you including the Amazon Fire devices on AOSP?

[Tester]

@ReadandBecomeStupid:

"So the fragmendroid situtation is even worse than imagined. This is a nightmare for the developers."

Admit it. Apple pays you to post this nonsense.

Actually, no, it isn't.
China is its own market, and so is India, and since most of the non-Google Androids are in these two markets it is really of no concern. Whoever wants to be present there needs to do a special version for the local language regardless of distributor.

But regardless of that: Smartphone app development is going down the drain anyway, as the quality of HTML5 support improves, there's increasingly less incentive to develop apps in the first place.

Most smartphone apps out there are nothing more than glorified website substitutes anyway. Doing this natively on PCs is not the norm and a few years down the line it also won't be the norm on smartphones.

[Per "wertigon" Ekström]

Apple is really paying very little attention to html 5 these days. So that is another way apple might face problems.

The more Apple refuse to play ball with the rest of the world, the more it will bite them in the ass later...

[Tester]

Apple obviously doesn't want HTML5 and web apps to succeed because it'd undermine their app store.

Hands up who is surprised here. I am not and this is precisely what I expected from a company that tries to impose their agenda on everything they touch.

What I see here is not fumbling around but deliberate sabotage.

[chithanh]

@ReadandLearn
> So the fragmendroid situtation is even worse than imagined. This is a nightmare for the developers.

It is quite manageable actually.

> Nope. Only nice amount of dividends.

So you are pumping Apple stock.

[Lullz]

Offline capable web apps are harder to manage for the end user. It's not just a single click to install them and another single click to uninstall. Also if the user clears the browser from all data, the installed web apps are also cleared.

[Catriona]

Just remember that Tomi predicted Hillary would win in a Reaganesque landslide. How did that turn out? The iPhone just turned 10 and Tomi didn't even acknowledge it. The single most important mobile device of all time deserves lots of attention right now.

[Abdul Muis]

@catriona

You need to differentiate the Job's iPhone and Cook's iPhone.

Cook's iPhone is a good device that build using the cheapest possible way. Kind of mercedes minus the luxury / expensive material.

[Tomi T Ahonen]

Hi Catriona

Hey, EVERY forecaster in EVERY industry makes mistakes. I am one of the rare ones who draws attention to every failed forecast as the forecasted issue comes to be. And I took clear ownership of the failed Hillary forecast - even as she did win more votes in that election. That type of comment - where I have published a whole blog article about my failed forecast - for you to make a point about it here, is pretty petty.

But if you're nitpicking on one Election 2016 related forecast that I got wrong (landslide, bearing in mind Hillary got more votes by a considerable 3 million votes) - will you be fair to point out how many 2016 forecasts I got right - I was among the first to forecast that Hillary would be the Democratic nominee (two years before election day), I was THE first to forecast Trump could win his nomination (mathematically possible) and among the first to forecast Trump would actually win the nomination. I predicted the outcomes of the conventions and the victors of each of the TV debates. And on election day, I had the correct forecast in terms of which candidate won more votes - but I was wrong on the SCALE of the victory (slim vs landslide). I trust Catriona you will be fair and honest with us, and admit, most of my forecasts relating to election cycle 2016 were correct while (sadly for me and happily for you) the final election result forecast was obviously wrong.

And again - every single forecaster makes errors in EVERY INDUSTRY. Including me long long before the Hillary forecast and I will continue to make errors. There is no such thing as professional forecaster who is always correct. That is impossible. But I am rare among professional forecasters to point out every time I turned out wrong.

But as to the iPhone, I was the first person who predicted the iPhone would transform not only the phone industry but OTHER industries like media, advertising and the PC industry. Said so in 2007. Forecasted it before a single unit of the iPhone was even sold. If you want to somehow draw from the failed Hillary forecast to any iPhone forecasts then YOU are drawing the wrong analogy. I have a near-perfect record of iPhone forecasts.

I was the first to forecast iPhone's regional sales pattern (correctly). I forecasted essentially every major evolution aspect of the iPhone (correctly) including stuff where I was a very tiny minority or alone, like MMS support and inward-facing 'selfie' camera for example. I was correct while avoiding stuff many of my peers thought Apple would do (correctly).

Far more importantly, I forecasted iPhone's market share peak (correctly) and the overall sales peak (correctly). There were times I was off in some of my forecasts by the timing but I never issued a forecast where I was wrong - such as I often asked Apple to give us a physical QWERTY slider/folder keyboard or a proper optical zoom camera - but I always made it clear, that while I wished those, I did not forecast those, and in fact, I forecasted Apple would NOT do those (correctly).

I forecasted accurately who are (Nokia, Samsung) and who are NOT (Blackberry) the main rivals to the iPhone. I forecasted accurately the needs of major changes to the product line such as large 'phablet' screen sizes - again, I was literally the first to do so and many iSheep on this blog argued forever how that was a stupid idea.

I forecasted the relevance of Apple's ancilliary products pretty well - the tablet side did better than I forecasted but even then, I was one of the few who said tablets and iPads would not overtake smartphones while many pundits then felt the tablet was the next big thing (it wasn't, I was more correct than most while Apple did better with the iPad than I originally forecasted). And the Apple Watch haha, you saw that, you were already here by then on the blog.

So in terms of Hillary forecasts and Apple - there is NOBODY on the planet who has gotten iPhone forecasts as good as me. Yeah, you're welcome. As to Apple's current 'plight' I was correct in predicting a slow slide path down the market share (as is now ongoing) heading towards 10% market share (give or take a point). Down from peak 21% to 15% now. NOBODY else said that when Apple was growing and hitting 21%. Nobody. You're welcome.

As to the 'unknown unknowns' to borrow from Donald Rumsfeld - who explained the Nokia collapse and how that would or would not help APPLE (and where, what regions?). NOBODY ELSE did that when Nokia announced its stunning surprise partnership with Microsoft. Not that I predicted correctly the extent of the largest collapse in any industry in corporate economic history of humankind (Nokia collapse, which only I did at the time) but I also indicated that Apple would pick up SOME of the spoils (correct, nobody else said this at the time) but that it would be Samsung (Apple's main rival) who would pick up MORE of the spoils (again.. nobody else said this, and yes, obviously I was 100% correct).

NOBODY else published that type of analysis in early 2011. NOBODY. And that was Apple iPhone's last hurrah, and it was a 'windfall' gain, meaning iPhone's organic growth had already stalled before Nokia fell, and obviously Apple's market share has now returned to the level if was prior to Nokia falling (ie 15%) from which it is now in slow decline. So even if there is a totally-unforseen change in the smartphone market - THIS BLOG and this expert has given the best forecasts about your favorite brand - better than ANY iSheep blogs you read. Isn't this so Catriona? This blog is BY FAR the most accurate when it comes to the major issues that impact the iPhone. From 2007 to 2017. Every single year, the best blog about iPhone's market performance related forecasts. Every. Single. Year. I trust Catriona you are honest enough to admit that, since you felt the need to bring up the Hillary forecast for which I already wrote a clear 'mea culpa' blog immediately upon the election result.

On the issue, is there cause to celebrate iPhone turning 10? Sure there is. Do you think this industry's unofficial historian might write about it? This you may want to hold your breath for, I might just do that. Especially as my ORIGINAL forecasts about the iPhone turned out darned good as well.

An article like that is not time-sensitive to exactly a day, where news like the UAE issuing the world's first passports to mobile is far more time-sensitive. Everybody knows this is the 10 year anniversary for the iPhone (that is not 'news'). So when I write my story, I will do it with care and the attention it deserves. It does not need to hit a particular exact date. And not publishing it yet, does not mean I would not want to celebrate that moment.

Tomi Ahonen :-)

[b]

@Tomi

What is your opinion on Google Home and Amazon Echo (and Homepod) type of devices and especially that they might allow in the near future voice calls? It looks to me like going back a little to phone land lines. Will be a fad or some niche for voice calls? Should be lumped into mobile devices?

[Tomi T Ahonen]

Hi b

Haha yeah, it is isn't it? The fixed landline phone is having a kind of rebirth. I hadn't thought of it from that angle, pretty funny actually.

As an evolution of the home, something like the voice-assistants do make a lot of sense for the 'connected home' and allow us to talk to our appliances and for various home-related services (including shopping) to be handled as conveniently as possible. Here the convenience of a mobile phone in our pocket is not always the best solution, so a 'central' home kind of universal interface makes plenty of sense. A bit like the ship's computer on Star Trek Enterprise that you could just talk to at any point in the ship, asking the computer to do something for you without walking to some touch-screen panel. (Probably some of the thinking into these solutions came as inspired by Star Trek Next Gen haha)

As to fad or real - I think very real. Most who get such a device will find plenty of use for it and an increasing utility out of it. The device has a good long life span (likely) similar to say TVs and soon they're about as common in homes as refrigerators, microwave ovens and vacuum cleaners haha..

Should they be lumped with mobile devices? It depends, some can/will be actually connected via the cellular network (hence counted as stationary 'mobile' devices like say a laptop with 3G/4G cellular modem, that only sits on the desk at home and is never moved out of the home). Many will connect via WiFi or possibly data cable - usually to a router on broadband but could also be WiFi router on cellular. If the end-connectivity is via WiFi then the actual device is 'not mobile' by my definition but in either case, as its not going to be moved about - it is a bit strange to call them 'mobile' devices haha. I think many IoT (Internet of Things) devices will fall into this category in the near future, where technically they connect via cellular but are built in and very stationary, so not 'mobile' as in portable/movable.

I have started to separate the machine connections from human connections in my mobile stats and we will increasingly see this dimension. It will mean 'inflated' ie large mobile user or subscriber numbers that will not be in relationship to humans alive (due to the IoT device universe often connecting also via cellular).

Tomi Ahonen :-)

[b]

@Tomi

I guess that the next few years the voice-assistant will be the hook which will draw people to such devices. As soon as these devices will be able to make/receive voice calls (via Skype, Wifi, SIM cards, etc.) to/from phone numbers will take off.
At least the old people (that do not have good vision anymore) will be drawn to this kind of devices for making/receiving phone calls and sending/receiving SMS/text messages (and get weather report, listen to radio, e.g. http://dailycaring.com/amazon-echo-for-dementia-technology-for-seniors/ )

[Tester]

Strange.

The people I talked to were just scoffing about the whole notion. I guess it requires a special mindset to buy into the whole idea and I cannot find that anywhere in regular people. That, btw, applies to IoT in general.

@b:

"At least the old people (that do not have good vision anymore) will be drawn to this kind of devices for making/receiving phone calls and sending/receiving SMS/text messages"

I've yet to see a single person above 60 who may be interested. Most are just scared about this newfangled tech they do not understand.

Well, we'll see. So far only some tech geeks seem to be sold on this stuff. As long as that stays it won't take off. It will only expand into the general public if there's some bona-fide advantages to be had and so far not much that has been said qualifies.

[Winter]

@Tester (and others)
I think you misunderstand the aim of these "units". They bring AI to the home. Apple and the rest want to become knowledge pedlers, instead of mere information libraries. They seem to see the best avenue to this goal is to set up an interlocutor or dialogue partner in the home.

The (distant) idea would be to have an AI conversation partner to talk to at home that can ultimately serve butler tasks.

We are far from that point, yet.

[Tester]

@Winter:

"They bring AI to the home"

Yes, and that's precisely where most people I discuss this with are calling it quits.
Say what you want but outside of tech geeks I absolutely cannot find a single person who is sold on this, and this goes through all age groups.

[Winter]

@Tester
No one cared what the geek in the street thought about railroads and the telephone. They were rolled out anyway.

[Tester]

And so was a lot of stuff that sounded cool on paper and never gained mainstream acceptance.