Do Androids Dream of Being iPads (or Why Many Android Tablets Are Sold But Not Used)
If you’ve been following Android news this past week then you’ve probably read about Google’s boast that over 70 million Android devices have been activated and that half of tablets sold this year ran Android.
This boast caused quite a few raised eyebrows this week as tech pundits tried to square the claim with the general knowledge that Android devices aren’t showing up in server logs or ad networks and the general belief that no one uses Android tablets.
I took an interest in that topic yesterday and I found an answer which surprised me.
Marco Arment tipped me to this line of inquiry, by first pointing me at a question raised by John Gruber of Daring Fireball:
I’m curious how Google squares these claims with all the usage share numbers that show Android tablets at far below 50 percent. Either the usage share numbers are wrong, or people just don’t use the Android tablets they buy.
Arment also tipped me to speculation by John Moltz:
I wonder if they include non-Google versions of “Android” in these numbers. But in answer to John’s query, if you were not entirely committed to tablet computing, wouldn’t you be likely to buy the cheapest tablet available? And when the user experience doesn’t wow you, you tend not to use it. It’s obviously not like that for everyone but I wonder if that doesn’t explain some of this.
If you are a contented Android user (like me) then you may be tempted to simply write this off as unfounded speculation. I was going to do just that when, but the next day a reader told me that his server logs showed that hardly anyone was visiting his site from an Android device:
On our web site (http://physicstoday.org) we have over 9000 visitors a month from tablets (growing at roughly 7% a year but that might change when the students head back to university). About 1000 of them are android users, the rest are using iPads as their device. That ratio has been consistent for the last year.
This data point is particularly interesting because it mirrors the info released by Chitaka, which reported that iPads represented 84% of the tablets on their ad network (in June as well as earlier months).
I would also have liked to cite Statcounter as a third independent source but I can’t get their report generator to work. So I am limited at the moment to just 2 data points that tend to disagree with Google’s boast.
One could take the easy road and simply ignore one side of the conflict or guess at a simple answer, but yesterday I instead took the slightly more difficult path.
I went looking for more data, and I found out that the situation actually isn’t much more complicated than consumers buying Android because it was cheap and then setting it aside because it was unpleasant to use.
It turns out that the app analytics firm Flurry has been posting relevant data on their blog for some time now that fills in the holes that some pundits are paving over with their assumptions. For example, in June Flurry reported that they’re tracking more Android devices (smartphone + tablets) than iThings:
As you can see, the latest data is from April 2013 and it shows that there are more Android devices in use than iThings.
At this point we have arrived at the question I originally set out to answer: why aren’t Android devices showing up online?
As you can see from the above chart, the devices exist and are in use. They’re just not being recorded as being online. And what’s even more interesting is that Flurry has data from April 2013 that shows this as well:
We know that Android accounts for half the devices in use, but Flurry shows that Android doesn’t account for half the time spent browsing. Now that is an interesting discrepancy, is it not?
I don’t know about you but it makes me wonder if perhaps the stock Android browser is so marginally adequate that users avoid it.
That might be causing Android users to spend proportionally more time in apps than in the web browser (as compared to iThing users). This could have been caused by the quirks of one platform pushing users one way while the other platform pushes users the other way.
But according to Flurry, Android users are actually spending less time in apps than iThing users:
So at this point we have reports that Android devices are being bought and they are being used, only they’re not being used as much as iThings.
I don’t know about you but I am beginning to suspect that there is truth to the idea that Android devices offer a poorer user experience than iThings. The data would seem to support that argument, don’t you think?
Please note that I am not saying that Android in general is worse; I love my Kindle Fire HD and I have found numerous imminently usable cheap Android tablets. But I do think it is possible that the average user experience, when you consider all Android devices, rates far below that of iThings.
If I had to guess I would bet that the open source aspect of Android is dragging down the average. Any fool can slap together an Android firmware and throw it on a device, no matter how crappy, and then sell it to an unsuspecting consumer. That is not a problem that Apple has to deal with.
It’s the dregs of the market that are dragging down the average, not the less rough budget Android devices like the Nexus 7, original Kindle Fire, and the like. If I am right then this is a situation that Android is going to have endure pretty much forever.
The problem of a poorer user experience is probably never going to go away, and that is because the devices that are dragging down the average never actually leave the market while more are added everyday. You can still find devices released in 2010 and 2011 on the market today as refurbs and even in new condition. (I still get traffic on the how-to posts covering those tablets.)
The original Pandigital Novel tablets (black and white) are still available and they are just as crappy to use today as they were then. The same can be said for the original Velocity Micro tablets, Viewsonic’s POS gTablet that was released in 2010, the many generic Chinese tablets released in the past 3 years, and even the crappy Coby tablets I reviewed last year.
If nothing else, one takeaway is that Android is trading quality for quantity. So the next time you read about some huge number of Android users, ask yourself how many are actually satisfied with their gadgets. Chances are that will be a much smaller number.
And another takeaway: The stock Android web browser is probably the worst part of the Android user experience. It should be used proportionally more than it is.
Syn July 27, 2013 um 1:30 pm
I have said this from the beginning. Google sacrifices quality to have those inflated numbers. Even across the the OEM’s that aren’t junk, (Asus, Samsung, etc) the experience varies from one device/brand to the next.
I have an iPad, Kindle Fire HD, and Galaxy Note 10.1. Because of all I can do and use the S-Pen for, I like it best. For near trouble free, user experience, I go to the iPad. Heck, I’m typing this on my iPad now. On my Note, typing this would be a nightmare because the keyboard skips letters as I type.
I don’t care that google gives any and everyone a license, but it would be nice if they held them to some form of quality.
Jordan July 27, 2013 um 1:57 pm
Or, most Android devices in use are smartphones, and many of these are cheaper phones. These users didn’t buy the phone because they really wanted to buy apps and use the internet, but because they were cheap- about the same price as a non-smart phone.
Tablets are a different animal than smartphones (right now at least). Moving your analysis of tablets to smartphones/tablets together can confuse the issue. You need to look at tablets in use and not just last years sales numbers or a tablet/smartphone aggregation- iPads have dominated up until the last couple years, and they still own the tablet market in a way they can’t say about smartphones anymore.
Becki July 29, 2013 um 6:05 pm
I was looking at the Flurry numbers, which state tablets + smart phones, and like you thought that should have been separated from the tablets vs iPad. We know that there are a lot of cheaper Android smart phones out there being used, more so than the cheaper Android tablets.
I think in some ways, the average person is still trying to figure out what a tablet should be used for, not just what it COULD be used for (reality versus advertisements). I knew what I wanted out of a tablet, and what its limitations are, and I’m very happy with my ASUS Transformer. Many people outside of the tech world were not as educated, and so probably bought them as the "in" tech and then didn’t know what to do with them.
(Note that I’m not putting myself into the "tech world" population; but I do a lot of research trying to keep up with the world of books for my job in a library, and tablets fall into the same area as ereaders for our purposes.)
Thomas July 27, 2013 um 2:10 pm
Flurry’s usage statistic ignore a lot of actual use. They only measure browsing on their networks and use of apps that report to them.
I use my tablet primarily as an ereader. I run FBReader (Open source with no Flurry spyware) with the wifi turned off to save power. I rarely browse on it as I have computers that do the job much better. As far as apps go, I’m particular about what I install, in order to avoid spyware. I doubt any of my usage is being reported to Flurry.
I don’t doubt that Android owners use their devices less than iOS users. If I’d spent that kind of money for a tablet, I’d be looking to use it whenever possible. But I don’t think the gap is as wide as these statistics show.
DebbyS July 27, 2013 um 9:40 pm
Pretty much the same for me, Thomas. I use mine to occasionally read a comic book (though at 7″, my Galaxy Tab 2 isn’t very good at that), or to watch an MP4 TV show (works fine for that). The device has not since I bought it seen a wi-fi signal of any kind, nor do I expect to go seeking one out. Extra aps I’ve added I have sideloaded. I imagine I represent a small percentage of Android users. I also have an Ectaco Jetbook Lite that I use a lot more for ereading.
Thomas July 27, 2013 um 11:25 pm
I probably turn my wifi on once a month or so. Every time, it wants to update 20 or more apps and takes about a hour to do so if I let it.
Part of the discrepancy in usage may be the form factors of the tablets. Apple sells many more 10 inch iPads than minis, whereas most Android tablets have 7 inch screens. Trying to browse the web on a 7 inch screen is really annoying.
Purple Lady July 27, 2013 um 2:20 pm
I don’t use the stock browser and when I use other browsers, I have the user agent set to desktop. I also use Adfree so I might not show up when I use apps either. I wonder what browsers they track and how they track them – there are a lot of third party browsers and most are much better than stock.
cookie July 27, 2013 um 3:15 pm
I don´t browse much on my tablets. I do pull down a lot of articles using gReader pro, so I doubt those get counted as Anroid devices.
Al the Great and Powerful July 27, 2013 um 4:15 pm
I use my Nexus 7 every day.
I read 4-6 hours a day, using Calibre Companion to manage my books and Moon Reader (except for Kindle files, which I read using the Kindle Reader). Does Flurry track that?
I use several map and navigation apps, access my photos on Dropbox, and use another Dropbox account for company images. Is Flurry mnoitoring my usage?
I play games (Carcassonne and Thumbzilla at the moment). Not sure they could tell.
I watch TV episodes and movies, loaded from my computer or streamed from a portable hard drive. No Flurry involvement at all.
When I’m writing I listen to Naturespace (I’m partial to rain and water sounds a soothing background noise). No idea what info they give Flurry.
While I shop, and look things up, I do not do much random internet surfing.
So that’s my Android experience, its as valid as anyone elses, and probably at least 66%* invisible to Flurry tracking.
*pulling a number from my fundament, in true flurry style.
Alexander Inglis July 27, 2013 um 5:01 pm
If we assume the conclusion is true: any given iPad is used more often than any given Android tablet, I would not doubt that. In my case, I have two Android tablets I use daily: a 3G connected Samsung Tab 10.1 and the wifi Nexus 7. While I do some Internet using these devices, that’s the least of the activities — certainly under 10 percent. And on the Samsung Galaxy S III, I virtually never use the Internet through a browser.
Yet, I am very satisfied with these devices and I use them in quite different ways. As Android’s installed base shifts to 4.x and up, you’ll find fewer people complaining about lousy apps. And none of the research ought to predict Android won’t have complete dominance of the category over time. Apple has always focussed on niche, premium products, quite satisfied with market share under 20% or even under 10%. Tablets will settle into this range over time, too.
yuzutea July 27, 2013 um 6:16 pm
Just because some Android app use is not detected by Flurry doesn’t mean anything, unless Flurry is somehow vastly more effective at detecting iOS app use. I’d think that some iOS app use wouldn’t be detected either.
Anyway, what would really be interesting is seeing the usage stats for more premium Android devices like Nexus 10, or Samsung’s phone flagships.
Thomas July 27, 2013 um 11:44 pm
It’s anecdotal, but from what I’ve seen, most iOS users add apps without a second thought as to whether or not they’re being tracked. They assume that anything that comes from the App Store is completely safe. Android users seem to be a little more picky about what permissions their apps want to use. With Android being an open platform, we have to keep security in mind when downloading.
I would like to see separate statistics for high-end Androids, but those might not be accurate anyway. The custom firmware I added to my cheap Polaroid tablet identifies it as a Samsung Galaxy 2 so as to get around some possible limitations.
Ron K Jeffries July 27, 2013 um 8:24 pm
With great respect (I love your stuff) Marco is a well-known and very smart Apple fan person. But his bias is mind boggling, IMHO.
I use Android on a phone and a 2012 Nexus 7. The Nexus 7, (last year version) has been a delight to use. Period. My wife selected an iPad Mini, it is a great device, but is not substantially better than a Nexus 7 tablet.
I don’t care enough to peel this onion apart, but it does not fit with my personal experience with Android. Watch the new 2013 Nexus 7 is better in many dimensions than the original, and happens to have a display with significantly higher screen resolution than iPad Mini.
The Android software varies in quality. You can find week apps, yes. but there’s a rich selection of solid and deep, with great user interface and thus provide a satisfying user experience.
My short, tl;dr summary: your position on this is dead wrong. 😉
Nate Hoffelder July 27, 2013 um 8:41 pm
Would you believe I went into this with the goal of finding data to prove Marco Arment wrong?
Seriously, I am an Android enthusiast and I wanted to show that Arment was completely wrong. Unfortunately that is not the data I found.
You should go back and read the post. I argued this based on data from Flurry, not anything written by that Apple snob. There are any number of reasons to dispute my conclusions but guilt by association is not one of them.
paul July 27, 2013 um 11:01 pm
As I provided Nate the physics today data, I want to make clear that I find our stats perplexing based on the number of android tablet users. Hence I’m just as interested over why there are so few hitting our site as I have to decide what products to invest in, based on this and surveys we conduct of our readers. I think Nate is right, its more apps and the poor quality of the cheap tablets turning people off.
Nate Hoffelder July 27, 2013 um 11:19 pm
Thank you for the data. That was indeed what inspired me to go look for an explanation.
paul July 27, 2013 um 11:03 pm
Plus the web browser does suck on it
Richard Adin July 28, 2013 um 6:48 am
I have a Nook HD+ (I also have the original Nook Tablet — before the HD designations) and I use it everyday for at least 3 hours and often for longer. What I don’t do is browse the Internet or use Apps. I primarily use the HD+ to read ebooks and occasionally to lookup more information about what I am reading. Also, occasionally, maybe once every 2 months, I watch something on Netflix.
If I want to browse the Internet or check my e-mail, I use my desktop computer. There is nothing particularly unusable or unfriendly about the access to mail and the internet on the Nook HD+, it is just that I prefer my desktop for those purposes. Besides, I am not a lover of Chrome and never use Gmail (or Google+ or any of the other Google products). I am accustomed to and prefer Outlook and IE, which are on my desktop.
When I will be traveling, however, such as my trip in September, I will use the HD+ to access e-mail and the internet, but until then, why bother when I have something I prefer right at hand. I suspect there are a lot of people like me.
The Ogre July 28, 2013 um 1:19 pm
"I don’t know about you but I am beginning to suspect that there is truth to the idea that Android devices offer a poorer user experience than iThings."
Probably because most developers seem to treat Android users as second-class citizens, especially if those devs are coming over from iOS (where they can be lazy and almost do one-size-fits-all coding, unlike on Android where the hardware configurations are almost as varied as the lap- and desktop world).
So they put out half-assed efforts, don’t see the sales figures they think they’re owed, scream "it’s all because of pirates!" (though copyright infringement rates on software are little different than on desktops), and go back to iOS (where the rates are actually still quite significant, but far less reported thanks to the Apple Reality Distortion Field).
The Ogre July 28, 2013 um 1:21 pm
"(though copyright infringement rates on software are little different than on desktops)"
Er, meant "(though copyright infringement rates on Android are little different than on Windows desktops)".
[Yay for revising the line 5 or 6 times ’til I lose track of things. :P]
Mike M July 28, 2013 um 11:54 pm
Instead of iThings, iPads and iPhones are actually referred to as iDevices.
Nate Hoffelder July 29, 2013 um 10:54 am
Somebody around here July 29, 2013 um 10:57 am
This only partly related to the author’s post, but I want to emphasize:
IGNORE ANALYST REPORTS
IDC, Gartner, Strategy Analytics, Asymco, Digitames (LOL) and the like are a scam. They don’t actually analyze anything, they MAKE UP the numbers. This was proven when Samsung released its tablet sales numbers: in 2012Q2, US sales of Galaxy Tabs only totalled 37.000. IDC had estimated a global number of 2.4 million, and the other firms consistently reported millions of Samsung tablets sold every quarter.
It’s all bullshit.
Now take this 70 million figure from Google. I don’t have the transcript, but I have seen two versions: that this figure is the number of Android tablets activated so far, and that it’s the number they expect to activate by year’s end. Let’s assume the former, and let’s also assume 30% of Android tablets are not counted by Google. That’s 100 million activations so far.
Well, the Strategy Analytics "report" for Q2 just came out. They have Android tablet shipments of 34.6 million…yeah, for last quarter alone.
Hold that thought. There have been 10 quarters of Android tablet sales (since Honeycomb). Or 4 quarters since the Nexus 7. Or 11 quarters since the Galaxy Tab. And this includes 1-3 holiday quarters when tablet sales explode.
And Strategy Analytics is saying last quarter alone had one third of all Android tablet sales. Or 50% of the sales reported by Google. Remember that the April-June period had no holiday sales, no Chinese New Year, and no significant tablet launches.
Also remember that 34.6 million is far, far higher than the numbers reported by Apple during holiday quarters – they have never reached 20 million. And presumably, Apple still holds a majority of the market in the US – these tablet have been sold…somewhere else.
Who the hell bought these 34.6 million things? And why don’t they show up ANYWHERE?
PS: I know your post used ad or web browsing analytics firms, rather than the pure analyst houses I mentioned above. Now, I would guess these are more reliable. But still, very often they only track devices in a specific part of the world. This part of the world usually is *not* China, Korea, Spain, India, Latin America, Africa, and so on. All of these are Android strongholds and Apple’s share of the market is much smaller than in the US.
So web browsing stats are useful, but they aren’t the be-all. And I imagine an Indian Android user browses far less than an American iPhone user.
Nate Hoffelder July 29, 2013 um 6:41 pm
It might not be relevant to this post but it is usually worth repeating whenever the topic turns to market statistics.
Syn July 31, 2013 um 12:59 am
I don’t think tablets shipped and tablets sold are the same thing, but I could be wrong!
Thomas July 31, 2013 um 2:05 am
>>>This was proven when Samsung released its tablet sales numbers: in 2012Q2, US sales of Galaxy Tabs only totalled 37.000. IDC had estimated a global number of 2.4 million, and the other firms consistently reported millions of Samsung tablets sold every quarter.<<<
This is false, the 37,000 figure was taken from the court records of the Apple-Samsung trial, and only included Samsung's first generation tablets. Samsung had discontinued those tablets earlier, and were selling the Galaxy Tab 2 series by 2012.
Ron K Jeffries July 29, 2013 um 6:28 pm
A friend (he did not request to be identified) provided these comments on this topic.
There are several explanations for why ad analytics firms like Flurry and random server logs show more iOS browser visits than Android. It’s very challenging (if not impossible) to answer these questions about large scale behavior without more data than the author has access to. However, you can bet that this will be the stat-of-the-month that iOS users bandy about to show they’re still "winning" some imagined, one-sided fight. A few possible explanations:
Stats collection is difficult. Both iOS and Android limit how much and what kind of information you can transmit off the device. Many popular applications have their permissions highly scrutinized, as excessive permissions can cause bad ratings in the Android market.
Android users simply may not browse the web as much, as an activity. I rarely do. I hit a few sites but not dozens or hundreds of random sites, it’s just not my internet consumption pattern. Some commenters hypothesize this is because there are many more small form-factor android tablets (5-8″) than large ones (9+"), and browsing the web on a small device is a pain still. I’m not sure if that’s the answer, but using web server traffic and in-app ads from different sources to build a combined view of user activity is going to be fraught with lurking variables.
Android tablets may just not get used as much – and I don’t know if that’s some grand indicator of anything other than mass market appeal. Since we know that Android devices are (on average) used less than iOS devices, primarily because of customer segmentation, it stands to reason a similar behavior may express itself on the tablets. I don’t exactly know why this is a bad thing – that’s like saying Ferraris are worse cars than big rigs because their owners drive them fewer miles and put less gas in them.
Technological differences – up until recently apps kicked to the background of iOS simply vanished, where on Android they reverted to a sleep state. In the case of games, this could close your game on iOS. So iOS users learned early on to simply leave things running in the foreground whereas Android users since day one have known they could hit the home button or turn the screen off without worrying. That difference in behavior could account for a tremendous discrepancy in "hours of use" for the kinds of apps Flurry tracks.