Quantcast: Android Eats into iOS Lead of Mobile Browsing

| News

The Android platform has been eating into the sizable lead Apple’s iOS platform has in mobile Web usage, according to analytics firm Quantcast. The company released new numbers Friday that show iOS with 56% of all mobile Web usage, with Android a distant second with 25% of mobile Web usage, but the delta between the two platforms is narrowing rapidly, as shown in the graph below.

According to Quantcast, iOS users did 68% of all the mobile Web browsing in My of 2009, whereas Android users did a mere 9% of mobile Web browsing. Since then, iOS’s share has steadily slipped, while Android’s has grown markedly.

Quantcast Chart

Source: Quantcast

Some of that share growth for Android came at the expense of Research in Motion’s BlackBerry platform, and some of it came from the other platforms (Nokia’s Symbian, Windows Mobilewhateveritscalledthismonth, and Palm’s Palm OS and WebOS). The rest, however, has come from iOS.

“In August Android took share from every corner of the market, putting in its best month share gain since November 2009,” Quantcast said in a blog post announcing the numbers.

The chart below shows the change in the various platforms’ share of mobile Web browsing month-over-month, quarter-over-quarter, and year-over-year.

Quantcast Chart

Source: Quantcast

Sign Up for the Newsletter

Join the TMO Express Daily Newsletter to get the latest Mac headlines in your e-mail every weekday.

18 Comments Leave Your Own

Deng

The “delta”. Really? Let me guess you, are you “rocking” a fine superphone? Yuck.

Bosco (Brad Hutchings)

There’s something else going on with this chart, but I don’t know what it is. On one hand, Android browser-share as percentage of whole is only up 2.5x since Nov-2009 (original Droid) while overall marketshare is up at least 8x as percentage of the whole. So G1 users apparently did a disproportionate amount of browsing. On the other hand, iOS browsing seems to continue falling with introduction of the iPad, even though it’s such an important browsing experience that everyone switched from Flash to HTML5.

Even if you believe Quantcast’s numbers or think they are particularly relevant, the bottom chart is strange if you want to demonstrate what is going on. iOS actually lost over 16% of it’s market share (1.0 - .56/.67) while Androids grew over 3x.

About the only explanation I can think of is that people hold onto their phones a long time and new users take awhile to get to the browser. So it’s another 6 months before Android passes iOS and probably 12 months before Flash on mobile shows up as relevant and Apple realizes they blew it by starting that war. Hmm…

Bryan Chaffin

The ?delta?. Really? Let me guess you, are you ?rocking? a fine superphone? Yuck.

I’ve no idea what you are talking about, mate.

Bryan Chaffin

Brad, iOS browsing isn’t declining, it’s share of total mobile browsing is. That said, that its share has declined only as much as it did when compared to Android’s (numerical) ascendency is a testament to the fact that iOS users do more browsing than Android users.

But that’s been known from the get-go, at least for iPhones. iPad browsing is likely only spiking the data further.

And, I’m up for another bet with you: Flash will never take off on Mobile. Ever. Let’s find a benchmark we can agree on and set the stakes. smile

Oh, and get me my essay from our last bet! smile

Eric Murphy

I’m guessing that the share of all web browsing done by mobile devices is increasing exponentially. So while iOS’s share of mobile web browsing may be declining, its share of total web browsing is probably increasing steadily.

That Android’s share is increasing faster would not change this fact.

Nemo

I tell you what is really wrong with Quantcast’s stats:  They appear to be false.  To begin with, the lower chart is so poorly labeled that one must make some assumptions to even be able to interpret it.  The lower chart says that it is comparing August 2010 stats, but the next panel is for what I assume are calendar quarters.  But that quarter can’t be the third quarter, which doesn’t end until 30 September 2010, so the chart must be comparing the second quarter of 2010.  Or is it?

The other major problem with both charts is that they contradicts the data that we have from Steve Jobs and Eric Schmidt.  For the second quarter, assuming arguendo that Mr. Schmidt was referring to new activations, Android was at 200,000 activations a day at some point in the second quarter.  Mr. Jobs now tells us in his latest presentation that iOS devices were never behind in new activations and at some point in the unfinished third quarter, iOS activations are at 230,000 new activations per day.  Well, since iOS devices have never been behind Android in activations and since the other competitors’ share of the market share for mobile devices has declined, losing share to iOS and Android, the only way that Android can be gaining share of web consumption, since Android has not gained device share relative to iOS, would be for iOS users to suddenly shift their habits so that they are engaging in dramatically less web consumption, an event which is so unlikely as to be incredible.

The other major problem with Qunatcast’s number is, if it is only looking at new activations how does it account for iOS devices huge installed base?  Or is it comparing, iOS huge installed base, nearly all of which is on contract, so that it can’t quickly change, with Android’s installed base, which is several times smaller.  If the later, and assuming that iOS users have not changed their web browsing habits, and since we know that Android has never activated more devices than iOS, how can you get the delta percentage changes in web consumption that is reflected in Quantcast’s charts, supra?  And if the former, you have the same problem.  Since Android has not gained share relative to the iOS and, if we assume that iOS users haven’t changed their amount of browsing consumption, Quantcast’s numbers don’t add up. 

Therefore, I don’t credit Quantcast’s stats.  Quantcast has some explaining to do, which can only credibly be done with it disclosing its raw data sets and the design and protocols of its sampling and analytical techniques.

Bosco (Brad Hutchings)

Nemo, I’ve got four iOS devices sitting in my drawer—3 Touches and a 3GS. I use an iPod Classic in the car because my (aging) Alpine doesn’t connect to iPod Touch. Most people I know who had a 1G Touch upgraded at some point. And a good percentage of iPhone owners upgrade to each new device as fast as they can. So while I agree with you that installed base bears something on these stats, it’s “active installed base” that does. I’m most shocked that the iPad is just propping up a declining trend. Presumably, they are counting sessions, and even without Flash, the iPad offers a browsing experience that’s more compelling in many ways than a phone. Browser on the phone is good in a pinch, and probably good for accessing media where you get to it and then passively absorb. But browser on a phone is not good for browsing for its own sake.

Nemo

Well, that is my point:  Quantcast number are just flat wrong, and thus, don’t reflect anything true, much less that the iPad result in a declining trend of iOS web consumption. 

And your latest comment, which is supported by other work, cast further doubt on Quantcast’s conclusions.  All, the studies that I’ve ready from survey organization of much greater reputations shows that most who own an iOS device, well north of 90%, plan to upgrade to another iOS device.  So the only explanation for such a decline in iOS web consumption is for users of iOS devices to suddenly start browsing less, which, of course, is patent nonsense.

So it doesn’t matter whether you are comparing the installed bases of Android and the iOS or the bases of newly activated devices—and of course any device recorded as being on the web must, must it not, be active?—Quantcast’s numbers don’t add up.

Bosco (Brad Hutchings)

I’m not sure about your 90% number, specifically with how plans play out at buy time. But I’ll grant it.

The bigger problem with these number is their relevance. I’d say iPhone/Touch are “90%” about apps. Android is “40%” about apps and “40%” about search (which includes Google apps like Navigation that are initiated by search), with the latter number increasing as phones move to FroYo. iPad is more like “50%” web.

Perhaps these numbers are an (un)interesting sign post, but they don’t really tell you what’s going on. Mobile is volatile, not only with market share but with actual usage of things. It’s definitely in discovery mode, putting a lot of processing and communications power in people’s hands, seeing what they do with it, and reacting. That’s why I’m bullish on Android. The platform and the philosophy guiding it are far more nimble than iPhone, BlackBerry, etc.

Nemo

Once again, users of iOS devices have been recorded as being the greatest consumers of both the web and apps.  That isn’t likely to have changed.  And while what I’ve read about the quality of the apps on Android’s MarketPlace leads to me to believe that Android is at best only 40% about apps, 40% about search, and 20% waiting on Flash to work, I doubt that Larry, Sergey, and Eric would feel much obliged to you for offering this concession about how poor an experience apps are on Android, both for users and developers.

Thus, as Sir Winston Churchill once said in response to Clement Attlee on the floor of the Commons, Android has much to be humble about.

Bosco (Brad Hutchings)

It’s funny to me that as an iPhone user, you see everything through the modal app filter. But I get it because I know Blackberry people who cannot survive without their BBM. LS&E are more than happy that their mobile platform is about search, because, are you ready for this, that’s what they came into this knowing how to monetize.

Contrast to Apple with the iPhone… First, they had to learn from the jailbreakers that native apps were even compelling. Then they get almost 200,000 apps at price points lower than they ever imagined with a higher percentage of free apps than they imagined. I was around and listening intently when the App Store launched, and I remember that their guidance was closer to $10 and $20 than free grin. So then to monetize that situation, Apple had to effectively kick all of the third party ad networks off the phone and set up its own PITA premium ad service that only big brands can get onto, and most who try complain bitterly about how backwards and restricting it is.

Apps aren’t a poor experience on Android. To me, they feel about the same as they did on iPhone. I just do more things on my N1 that don’t involve modally switching into an app than I ever did on iPhone. And judging from the sites and forums, I believe the experience is pretty common.

geoduck

What I find most interesting (assuming that the data is trustworthy) is the tiny share allocated to RIM. The large BlackBerry installed base notwithstanding, the users don’t appear to be surfing the web. Having the misfortune of using a BlackBerry I can tell you from experience that the browser is simply bloody awful. Terrible to the point of being nearly unusable. This is one of the most fundamental weaknesses of RIM.

Gareth Harris

Careful here. You have to be careful when comparing percentages rather than quantities. Often percentages get inflated from modest changes to small numbers. And when the total quantity itself is changing, some sector hay show a percentage decline while still showing great growth in units.

davebarnes

OMG!
Short AAPL.
Buy an Android phone so you won’t be left behind.
Buy a Chrome OS tablet and ditch your loser iPad.

John Molloy

These figures do not include the iPad. Part of the drop can be explained by the people using the iPad instead of the iPhone or iPod Touch. I have no idea why they left out that part of the equation… it couldn’t possibly have anything to do with making Android look spectacular…

Nemo

The problem is that Quantcast’s work is just incompetent.  The only plausible explanation for there numbers is that they looked at a very small sample of the trend in the activations of leading smartphone platforms, ignoring the installed bases of those platforms.  For example, even though Apple is activating substantially more phones, but its share of the total new activation is around 56% instead of its installed base of better than 70% for all of its iOS devices.

The problem, however, with that are at least two things:  (1) Quantcast does not tell us what its doing, nor does provide the raw data and primary stats that prove its results; and (2)  sampling over such a small period of time, such as parts of the second and/or third quarter(s), is worthless for extrapolating anything about long terms trends in either web usage and/or unit market share.  It’s like my sampling summer temperatures in Georgia on the coldest day of the summer and concluding from that the average summer temperatures in Georgia is 65 degree Fahrenheit.  And, of course, anything that dishonest is designed merely to make Android appear to bigger in the market and with a greater trend line than the statistical evidence will support.

In any event, anytime that statistical analysis is so badly done that one is left guessing what the statistician did and how he did it, the analysis isn’t worth the paper or bits that it’s written on.

Voice

This is an example of where absolute numbers (bandwidth consumed) would have made for a more informative chart than relative numbers (percent of bandwidth consumed) does.

Android has become a fairly successful and popular platform.  It has a competent browser.  Of course it’s going to be picking up a percentage of the total bandwidth consumed as its sales increase.  No surprise there.

Even including the percentage of ‘units’ doing the browsing for each platform would have helped convert these values into something informative though.

WARNING: Numbers in the next section are for example purposes only!

For example, if iOS devices make up 60% of the devices measured here, and Android makes up 20%, then Android is responsible for slightly more bandwidth consumption per unit than iOS.  If, on the other hand, iOS devices account for 50% of the devices and Android accounts for 30%, then iOS devices are responsible for more bandwidth consumption.

Without those numbers, it’s difficult, if not impossible to glean any useful information from the data presented.

Side Note:  Why do I sometimes get a CAPTCHA error.  I’ve never seen a captcha on this page.

Nemo

Dear Voice:  We are in accord on this point:  As presented, Quantcast’s stats are useless.  The iPhone 4 wasn’t even on the market until 24 June 2010.  Thus, the second quarter only had six more days, and of course, many iPhone customers were delaying purchase until the iPhone 4 was available.  And that was only one of several factors that would have Quantcast’s small sample of the period that it sampled so atypical as to be worthless for almost any purpose of comparison. 

Once again, if Quantcast was trying to do good work, it failed miserably.  If it was acting as a shill for Android, perhaps it succeeded, because people often accept stats like this at face value.  But as Samuel Clements once said:  “There’s liars, there’s damn liars, and then there’s statisticians.”

Log-in to comment