Validating the Android engagement paradox

Following yesterday’s IBM data, Monetate released a new study showing similar data related to retail browsing but covering a period of dates from Q3 2011 to Q3 2012.

This data also shows an acceleration of mobile shopping, from 7.7% of online in Q3 2011 to 18.8% in Q3 2012.

It also shows tablets growing to take about half of mobile traffic in a very short time frame.

The data also shows the iPad taking the vast bulk of traffic among tablets (88.9% vs. 88.3% from IBM).

The data also shows the iPhone taking 61% of phone traffic (67% from IBM) with the rest being Android.

The data shows the share of iOS being 74% in Q3 while IBM reports 77% for Black Friday.

The data also shows a very small participation from Kindle and Android tablets.

Finally, by using the install base data from ComScore, the data shows an absolute and relative decline in Android engagement. Android phones went from 0.47 of iPhone engagement to 0.4 in one year.

If you’d like to learn more about the Asymco method, come to Asymconf. The next Asymconf will be at the end of January near San Jose. Not only will you learn about how mobile is disrupting computing, you will also get to hobnob with the likes of Om Malik, Jean-Louis Gassée and Michael Lopp.


  • KitFR

    Monetate provides software as a service (SaaS) for various retailers. While this looks to be excellent data, it is worth keeping in mind that it almost certainly does not include Amazon. Consequently, the Kindle might be proving more of a success than this data suggests.

    • Kindle might be proving more of a success at selling Amazon merchandise.

      • Horace: if amazon sells tablet at break even or worse to make up on sales
        It seems this data shows failure based on number of tablets sold by amazon?

      • If neither IBM nor Monetate sees Amazon site traffic, as someone proposed, then the Kindle stats may be very distorted if Kindle users mostly buy stuff from Amazon and not from elsewhere, which is at least semi-plausible.

        I don’t know where either dataset is tapped from. If it’s tapped at the commerce sites, then it’s selective and may indeed not include places that don’t outsource their analytics. If it’s picked up somewhere else (such as ISP headends or peering points, which would mean the ISPs are the analytics customers), then it may be more broad-based and more accurate.

      • Very plausible, and I think very likely. But I don’t know if it matters.

      • oases

        At least semi-plausible? I would say at most semi-plausible. Kindle/Amazon usage is nothing more than a theory. I think it’s just a poor man’s iPad. Even Amazon has advertised it against the iPad.

      • iPad might be proving more of a success at selling Amazon merchandise.

    • Yesterday I considered this a problem too. But Kindle Fire tablets don’t have special access to a members-only section of the Amazon store.

      Monetate/Comscore/IBM doesn’t know if I’m shopping on Amazon using iPad or iPhone either.

      We don’t have perfect data but that doesn’t invalidate the good data we do have.

      • Does the Kindle Fire access the Amazon store through the web or does it have an app whose performance isn’t measured by ComScore et al?

        I have an app for my own web site which is still under test. I send and receive data for it using XML, bypassing the web site proper. I suspect Kindle might work in the same way, and that would mean ComScore simply would not record it as a visit or transaction.

        If that is the case here, we know nothing at all about Kindle Fire visits to the Amazon store.

        On the other hand, Amazon would probably trumpet the data proudly if it was highly successful, and as far as I know they have been pretty quiet.


      • comScore maintains a group of users who have monitoring software (with brands including PermissionResearch, OpinionSquare and VoiceFive Networks)[15] installed on their computers.[citation needed] In exchange for joining the comScore research panels, users are presented with various benefits, including computer security software, Internet data storage, virus scanning and chances to win cash or prizes.

      • Youre right, and I bet we know nothing at all. Anything we do know is potentially suspect unless it is information extracted publicly available data and therefore falsifiable.

        As far as I know Kindle’s Silk browser routes all traffic through Amazon servers. They do this to analyze, cache, refresh and prefetch pages. Ostensibly this is to speed things up, but of course the side effect is very broad analytics about user behavior regardless of where they browse to. It’s a marketers dream: spyware that improves performance, doesn’t leak to 3rd parties, and doesn’t tweak user incentives.

        Amazon is very and hyper-competitive. I doubt that anyone but Amazon has full-spectrum-analytics data on user behavior inside the site. Analytics firms can only do so much, especially without distorting behavior.

      • FalKirk

        “… Amazon would probably trumpet the data proudly if it was highly successful…” – David H Dennis

        If they had good news, they would be shouting it from the mountain tops.

      • Lee Penick

        If they actually put some “E” in their P/E? (All “P” and no “E”)

        I like Amazon but feel sorry for them.

        Look at “” how is Amazon ever going to make any kind of reasonable profit with competition like that?

        P/E compression is about as fun for an investor as a endoscope…there is gonna be a burr on their scope.

      • The question here is exactly what the raw data is. if they’re looking only at stats extracted from a web server or tracking bugs on the pages, you get one set of data (or two, since you can track clients by either cookie/userID or IP address). If it’s from network traces, what you see depends on whether it’s looking at just standard HTTP/HTTPS TCP sessions (ports 80/443), all TCP sessions by IP address, or all traffic by IP address. There are a lot of nuances in the collection process.

        Since they’re seeing browser platform information, I’d guess that the data is most likely gathered at the web server, where it’s easiest to get at that information, though it can be obtained from traffic traces as well, though less reliably.

      • Wikipedia says it was initially through a proxy server, but they changed that when it became too slow.

        Data from my web app is transferred via HTTP and I’m sure that’s true of Amazon’s too. So that could be valid if the equivalent to a proxy server was used. If the app just spies on all HTTP data and credits it to the platforms, then even an app that used XML would be counted.

        But if, as I suspect, the tracking is done via JavaScript, we aren’t capturing any app data at all.


  • It has to be the platform.

    Given the big numbers, the subsidies and the different price points of ios devices, medium user capabalities should be almost the same between platforms. I mean both technical skill and economic capacity of users, with this numbers, they are more or less the same or if different not so much different as usage numbers are.

    Question is: what is the real difference in the two platforms that imply this incredible different user behavior?

    God is in particular and in commerce a lot of small thing has to be got right to sell, but this is something new, the product has been already sold and yet there is something different that makes one product to be used for further purchases and the other one not.

    User interface, curated marketplace, confidence in the brand we can make our bet, but much analysis has to be done to really understand what is going on.

    In any case apple has created the holy grail of commerce, a reliable and usable point of sale. That is what I call a strategic advantage that is not easy to close for competitors.

    • I’m glad you see it. We need to move the ball forward here – some of the discussion the past few days feels a bit like the noise right before the US election: lots of data, all of it pointing to the same thing, and a lot of people arguing over why it was right/wrong. Lots of emotion, lots of opinion, not much science.

      As far as I’m concerned the preponderance of evidence points to a simple, falsifiable conclusion:

      iPhone and Android are hired to do very different jobs – and this is BECAUSE they are paid differently.

      • Not so much differently, given the subsidies the highest price android models cost like the iPhone 5 and the lower price cost like the iphone 4s or the iphone 4. They have the same plans and the same initial cost.

      • Looking at the Verizon site, every iPhone from the 8gb 4 to the 64gb 5 is subsidized $450 by the carrier.

        The only Android I found that receives a $450 paycheck is the Droid RAZR MAXX. The subsidies drop as the price of the phone drops, and varies between phones.

        They are paid quite differently.

      • You imply that ios is for high budget users and android for low budget. I don’t think this is true, in your examples you have an iphone at 0 price point and one at 200, that is a price point for every budget.
        Furthermore it is not true that low budget users are the one that buy less online, given the attention they put on their money they are more likely to scan online for discounted offers, they have less money to spend but they are a lot more so the final number should be considerable and it is not.
        In any case the use cases, the jobs, for the phones are the same whatever budget you have when you buy it, instead web browsing (that has no cost) is far more used by ios devices than android.
        It is in the platform, must be.

      • How am I implying that?

        Carriers pay iPhone differently from Android because iPhone does a different job.

      • It is a tautology.

        First you say:
        “iPhone and Android are hired to do very different jobs – and this is BECAUSE they are paid differently”

        than you say:
        “Carriers pay iPhone differently from Android because iPhone does a different job”

        There is not a cause, only a circle.
        In any case the capabilities of the devices are more or less the same, apps are still better for ios, but android is not that bad, construction quality is different but doesn’t affect jobs execution, only total cost ownership, so that choosy users can see the iphone cost as lower.
        Where do you see the iphone capable of doing a different job with respect to android?
        It can do the same job better, but different means that android is not able to do it.
        There are a lot of things iphone does better, from apps to services, and some thing android does better, I don’t think there are a lot of things iOS does that android in some way can not do.
        I mean, certainly there are but I don’t think they are the factor that imply this difference of use between platforms.

        There are a lot of android users that replaced their feature phone with android but won’t use smartphone features, but also a lot of iphone users that bought it for the aura without knowing its features or have prior use cases to be solved.
        But after purchase iphone users start to discover new uses for the thing instead of only phone calls and messages, while android users do not.

      • I’m in the car, ill try to break down my premises and conclusion as soon as I’m back at the office.

        I appreciate the debate, and I agree with your argument that the platform is what sets iPhone apart.

        I’m just trying to get to a very well defined theory of what that platform consists of and how it interfaces with the entire value network.

      • JohnDoey

        You can buy a $299 Android phone, but it will be a cheaper phone than the $299 iPhone.

      • Relentlessfocus

        I’m wondering if the aggregate term “Android” is too coarse a measure. I’m curious if phones with competing OSs of similar price ranges have similar use patterns and what skews Android numbers are the ultracheap/BOGOF etc phones that go to people with little interest in a phone?

        I’m also wondering if these comparisons hold up internationally.

      • What phones you suggest comparing?

      • Relentlessfocus

        I’m not that familiar with what’s out there in the US market but for example what are the use characteristics for a Galaxy SIII and iPhone 5?

      • JohnDoey

        Galaxy SIII is not a typical Android phone, that is an iPhone clone, it is the high-end option at stores that do not carry Apple yet (most of the world market.) Most Android phones are feature phone replacements, not iPhone replacements.

      • JohnDoey

        Yes. Android is just an open source project that device makers can use. Same as Apple WebKit. WebKit is actually more popular and more open than Android.

        Consider all WebKit phones. Why aren’t they all being used like iPhones? Why would you expect them to be? Some are iPhone-like, some are not. Has nothing to do with whether they use WebKit or Android.

        There are a lot of MP3 players, and only one iPod. There are lots of phones with calls, SMS, Web, and Java applets. Only one iPhone. It has what other phones have plus a free Mac and free iPod. Real native C/C++ apps built on OS X frameworks and real iTunes+iPod audio video. Those are very, very valuable things to get in a $99 device that also does all the generic phone things. iPhone is a unique, non-generic device that creates demand by examining itself clearly and then committing to being better every year so that the user’s commitment to iPhone is rewarded. Nobody else is doing anything like that. Nobody else should be expecting similar results.

      • Space Gorilla

        I wonder if the difference can be explained by viewing Android as a smartphone platform and iOS as a mobile computing platform. The details may matter. One is being used as a smartphone and the other is being used as a mobile computer (the devices are hired to do different jobs and one job, iOS, involves much more usage). It remains to be seen if Android can ever get beyond being the new default smartphone platform, and while a smartphone can do many things that we would call computing, it is not quite on the level of a true mobile computer.

  • Tom Davis

    Invite Brian S Hall. He would add a little spice to the proceedings.

    • normm

      He’d be a distraction. He isn’t in the same league as the others in terms of insight or logic.

  • Bruce_Mc

    I suspect that a lot of people in the US do not choose to buy an Android phone. They basically let the sales person at the cell phone store choose their phone. Perhaps all they hire their phone to do is phone calls, texts, and email. An Android phone will do that, and it’s the same price to the customer (free) as a feature phone.

    This hypothesis can be tested by walking into a carrier store. Tell the sales person you have an old feature phone and ask them what kind of phone you should get. How hard do you have to work to get the sales person to recommend an iPhone?

    • Or how hard do you have to work to get them to sell you a feature phone again, that doesn’t have a mandatory, expensive data plan attached?

      • Bruce_Mc

        Very good point. Another thought: where are the feature phones with 3.5 inch (or larger) screens? If the job to be done is reading email and texts easily, a feature phone with a tiny screen is not going to be attractive.

  • johnambani

    Talking about subsidies on phones…what is the average payback to a carrier for an iPhone subsidy (or better yet what is the IRR on that customer) vs low-end and high-end Android phones? Clearly it seems that data usage is much more substantial among iPhone users but I don’t know the specific data on that.

  • Marc

    Simple: Android picked up the market for featurephones while iOS captured the market intended for smarphones. I know may people in Europe that have an Android phone without a data subscription. These are the orphans of Nokia, that now get a droid at the same price point…

    • That should be quantifiable if the European carriers report data subscriptions separately from voice — plot the data subs against smartphone sales. I don’t know if any of them break things out that way, though.

  • randomness9090

    Is there, anywhere, a real study that breaks down the marketshare or user base of “Android phones” by type of phone? All we ever hear about is the massive marketshare of Android, and my impression is that Android fans generally think, most likely inaccurately, that this marketshare is composed mostly of high-end Samsung or other iPhone-like smartphones.

    I have to admit that I find it frustrating that journalism is so simplistic as to never dig into this question. Most articles never get past the “Android is winning” headline.

    • Presumably many of the market research companies gather this kind of data, but they don’t seem to be making it public. I assume this is the kind of competitive data they sell back to the phone vendors for high prices, so it’s not in their interest to make it freely available.

      Google probably has this data, at least for devices that activate via Google. (I can’t imagine that Google doesn’t collect every scrap of data it can about activations.) But they don’t publish it, either. And a lot of the low-end “Android” phones out there don’t actually activate via Google anyway, since they’re not including Google apps.

      The closest stand-in I can think of for this data is Google’s public information on the relative market shares of different Android versions, which I suspect correlate at least moderately well to the low-end/high-end phone division — only the newest, high-end devices seem to run the newer versions of Android, from what I’ve seen in the blogosphere. This is probably driven by software performance and memory requirements which drive processor and memory component selection, and thus increase costs. Software requirements are like taxes: they tend to only increase with time.

      • El Aura

        Google publishes data in screen size categories and by OS version (and anything still shipping with 2.3.x probably ain’t the high end models).

    • Such a study cannot exist because the data is not public and cannot be derived. Even knowing the total number of Android devices sold or in use is a matter of conjecture. (Journalism is not analysis.)

  • robin

    There have been a number of reports that the Android tablet share is much higher than these numbers would suggest:

    In the past 2 years, I’ve seen one Android tablet in the wild, but I see iPads all the time, so this seems very hard to put into perspective. Where are all these devices being sold?

    • neutrino23

      Same here. I travel a lot and the airplanes are full of iPads and a few kindles but I rarely see an android tablet.

  • Claude Hénault

    Here, off the wall, is my thesis, which I have no means of checking.

    People have more sensitive antennae than they are credited with having. When they are being used, they know it, and down at the barely conscious level, they mostly resent it.

    At the level of a meme, people are aware that when they get an Android device, they belong to it, not vice versa. When they use it to cruise the web they are aware they are peddling their ass for Google, and it grates that they are being used. When they get an iDevice they are aware of the reverse: it belongs to them. The deal with Apple may have had excessive margins attached, but at least that’s the end of it, as far as exploitation goes.

    And I think Apple is counting on this and plans to increasingly market as the privacy phone, privacy pad, privacy pod. I believe Apple’s doppelgänger patent of last June, with its capability to protect true identities by scattering false ones to the web, shows an intent to clearly demarcate on the privacy front. Their painful shedding of Google Maps and Beta release of Siri are further clues they have identified privacy as a marketing wedge that they are, and will increasingly, exploit.

    • Doppelgänger patent? Link please!

    • Nick

      I have been under the impression that people are not aware of how the Google-Android devices’ strategy are like, and in my experience they are not considered as common knowledge even among Android users.

    • Jerome

      “When they get an iDevice they are aware of the reverse: it belongs to them.” For me, this is absolutely true. I perceive the phone as something I own and paid for, and the same holds true for the apps I use. I am the customer. And this makes up for a big difference, because the one who is paying is the one everybody caters to.

      On Android, this is far more blurry. More often than not, the user is NOT the customer, but the product being sold. Apps and phones tend to hover around the bare minimum, the lowest common denominator that the user will put up with. Simply because the user, for most participants of the Android ecosystem, is not the customer. The user is not the one who must be satisfied, it’s enough to aim for the pain threshold of the average user.

      If you want me to pay for your app, you have to do something right and meet my expectation if the app is going to be hired in return for money. If it’s for free, the user will put up with a lot of crap – but in the end, it’s not a pleasant experience and the user hunts for the next applicant that promises to do the job mediocre or worse, but “for free”.

      My theory is, that the lesser engagement is a long term consequence of the Android ecosystem. If the user is not the customer, the user satisfaction is an afterthought instead of the highest priority. As a result, the user doesn’t use it as much.