The smartphone wars, one year later
The iPhone leads the pack, Android is gaining, everybody else is losing share
It's been a year since Google (GOOG) released Android OS, the open-source smartphone operating system widely perceived as the most likely to overtake Apple's (AAPL) iPhone in the long run.
As it happens, Google this month also purchased AdMob, the world's largest purveyor of mobile phone advertising. So this seemed as good a time as any to take a snapshot of the changing smartphone marketplace, as measured by ad requests to AdMob's network.
We reviewed a year's worth of AdMob data — including the October numbers released Monday — and charted it on the graph at right (reproduced full-size below the fold).
There's a bias in the data, since AdMob ads run better on iPhone OS and Android devices than on, say, Research in Motion (RIMM) BlackBerries. But the trends are clear.
Over the past year, Nokia's (NOK) Symbian has lost the largest raw market share, down to 25% last month from 59% the same month a year earlier. In percentage terms, Windows Mobile (MSFT) is the biggest loser, down 70% in 12 months, with Symbian, Palm's (PALM) Web OS and BlackBerry OS close behind.
These numbers are based on worldwide ad requests. Apple's lead is even greater when AdMob zeroes in on the U.S. and U.K. markets. For a look at how the iPhone's share of the U.S. and worldwide markets have grown, see the chart prepared by MacRumors' Erik Slivka here.
Below the fold: A full-size fever chart of AdMob's worldwide data for all the major smartphone operating systems.
[Follow Philip Elmer-DeWitt on Twitter @philiped]
I live in Holland (Europe), do own a HTC Hero with Android, am online all the time and use it intensively. I have lots of apps installed and do use them. But I have never seen even one Ad from Admob. What do I do wrong? Why am I not in these stats? I feel ignored ROFL.
@Harry,
Elmer-DeWitt's blog is about Apple. He is not hiding any bias, it's right there in the masthead.
The graphs record your smartphone use even if it's over WIFI. It's not specific to 3G networks.
Nowhere did PED say the graphs were of marketshare. They are clearly of ad requests, and PED noted possible biases.
Once again, the charts are not of marketshare, but are a proxy for 'net usage, which is different. The reason being that not all smartphones are created equal. Apparently some are smarter than others. David Pogue has recently called the iPhone and Android phones like the Droid, App Phones, to differentiate from Smartphones.
As for Apple, I'm sure they will gladly tell you that market analysts like IDC, Gartner, NPD have measured their smartphone share at about 18 to 20%. And, I'm sure Canalys, whom Nokia uses to compile their data, will tell you that Nokia has gone from 69% smartphone share to about 40%, in just over 2 years.
As for Nokia, seeing as you know so much about them, did you know that the N900 uses Linux Maemo and not Symbian?
I agree with Jay,
All CNN articles, especially by Phillip Elmer are Apple biased only. Again Admob requests doesn't give the clear picture.
I myself have a smartphone, where i don't use a 3g data,
but only use pay as you go,only when utmost necessary, else I use only wi-fi. So i don't browse the web as much as iphone users. But that doesn't mean i don't have a smartphone.
So marketshare can't be compared to just admob graphs.
I know my friends in India and China, who have millions of Symbian smartphones, but don't access the web on their phones as much as we Americans do. So they have millions of smartphones, which don't get portrayed through such graphs.
Apple certainly has gained market share, but these kind of graphs are more mis-leading and made for manipulation of stock market, rather than depicting the real picture.
The real picture, which even Apple knows is that Symbian is still the leading platform. Does Apple report their smartphone share in their earnings? Nokia does rpeort their own estimated share in their earnings.
This is especially when 98% of American don't know about Nokia. I am waiting for Nokia to step up their game in USA in 2010, which will happen with Symbian^2 and AT&T partnership soon.
Has anyone even heard of Nokia N900?
Hi Philip. As you say the ads work best on the iPHhne – you didnt say how well or badly they perform on Symbian phones. Given these qualifications to the data – have you any idea how far off the figures and the conclusions are?
ex ped: Everything I know about AdMob's bias I learned from Dan Frommer here.
I feel strongly that Rim could be out of the smartphone market pretty soon being snowed under by competition.
This only confirms that the Blackberry is going the way of the Motorola Razr which I owned. I switched from the Razr to the Blackberry which was a game-changer in its day. I now own an iPhone another game-changer. The difference is that I can't see how I could ever switch from the iPhone because off applications like Dictionary.com and NeuroMobile that I rely on. The applications advantage makes the underlying handset technology less relevant. Apple now just has to keep pace with the handset technology and keep churning out great applications to maintain these positive trends.
"Apple will make lots of money…"
That's all that matters as long as they keep delivering quality products.
One last thought:
It is important to remember that AdMob records ad requests from the Mobile Web and dedicated Apps. The implications are that devices that only go to the Mobile Web will have their ad requests completely recorded.
Those devices that don't go to the Mobile web but go to real webpages will have their data underrepresented. Likewise, before developers started to embed AdMob ads in their apps, those devices that had lots of Apps, an alternative to the web, were also being underrepresented. Even now, since not all apps have AdMob ads, the devices that offer a plethora of apps are probably being underrepresented.
I would point out that feature phones still represent 56% of all ad requests as measured by AdMob. Does anyone really think feature phones, in toto, access the web more than smartphones? I don't. I think this stat clearly reflects the bias towards measuring tools. I don't recall what the term is in science, but there is term for that kind of bias.
@Rick in Nashville,
While that is true, the drop did NOT occur in the Spring, it occurred mostly in the Summer, in June, July and August.
Second, just because Apple is selling fewer iPhones as people wait for the new ones to launch, doesn't mean that existing iPhones aren't being used as normal. AdMob's smartphone request chart is not a snapshot of sales at any given time, they are a proxy for marketshare usage. So, installed base matters, and ease of access to the net matters. People waiting for new iPhones is only going to have a small affect upon that, certainly not enough to drop share 10%. Think about it, no other platform outside of Symbian even had 10% share. So, the amount of share loss was larger than RIM's complete market share, larger than WinMo's complete market share. Makes no sense.
This is interesting. I'd still pay more attention to Apple's quarterly reports. And that of other smart phone makers. Those numbers are less debatable.
We are seeing things change at lightning pace in the handheld arena. The iPhone just came out a little over two years ago. The app store is only a little over a year old. Look how much the world has changed in that short time. It is far too early to say how this will turn out. For the short term it does appear that RIM will muddle through, Apple and Android will grow quickly and the other platforms will lose share.
Apple will make lots of money, but will never grab majority market share in anything
You know, making lots of money tends to be what businesses want. Market share is what pundits want.
The reason why Apple's share drops in the spring is that Apple has historically introduced a new iPhone in the June/July time period. Buyers know that and Apple typically tightens production and inventory during the period leading up to a new product launch.
"Apple will make lots of money, but will never grab majority market share in anything."
Hmm, except for mobile media players, digital music downloads and a mobile app store that's ten times bigger than its closest competitor.
I hate to add more, but looking further at the odd decline in iPhone ad request share of 10% from May to August, it can't easily be explained by an increase in Palm share by 3% or Android share by 4%, as the other players, like Symbian, RIM, and even WinMo appear flat in the same timeframe. How is it that those laggards have maintained their share in that timeframe when the iPhone loses 10%? Seems completely nonsensical. Reason tells me that while launches of Palm and Android around the world could grow their share faster than the iPhone, the fact that iPhone loses in comparison to Symbian, RIM and WinMo makes NO sense.
If I were working at AdMob, that would have jumped out at me, and would be something to analyse and discuss in their report.
ex ped: Good questions, Ken. If you're interested, there are links to AdMob's archives here.
A few comments:
One, share is confusing, when the overall market is growing quite fast. Why does the iPhone share drop 10% from May to August? Doesn't seem to make sense. In absolute terms sales of iPhones are going up, users are using the phone just as much as ever, so the only way to explain it is that other phones are requesting at a faster rate of growth. The only 2 mfrs that have growth in this period are Android and Palm, each growing what looks like 3 to 4%. Palm goes from 1% to 4%, while Android goes from 4% to 8%. So, there's 7% of the 10% explained, but not explained completely. An overlay of handset intros would help. Presumably, there are Android handsets being intro'd around the world, and I'm guessing the Palm Pre was intro'd in the UK during this time.
A chart with absolute numbers of requests would be useful next to this one showing share, as it would help provide a little context.
Two, it would be good to note that Smartphone ad requests still only represent 44% of all ad requests that Admob recorded. So, those other feature phones are still serving up the majority, 56%. Hard to believe.
Third, it would be nice to know to what percentage of apps, AdMob serves ads to. What I mean, is if AdMob serves ads to 1000 apps, regardless of whether it's the 100,000+ AppStore or the 12,000+ AndroidStore, then there would be bias towards the AndroidStore, as AdMob would be serving ads to a far higher percentage of apps in the AndroidStore than the AppStore. Just food for thought.
Four, and before anyone starts spewing FUD about the iPod touch skewing the numbers, I would point out that when AdMob specifically states Smartphone, they are excluding the touch. In other mobile aggregate data they do include the touch, but not in their Smartphone only data.
Only "cheapsters" can rule the world?
What do you call the $99 iPhone 3G (w/ 2 year contract)?
Many lesser phones have been offered for that price in the past. Certainly not rock bottom pricing, or anything like the recent "buy one, get one free" promotions on Blackberries, but certainly not expensive, at least in the US.
Will anyone in the world offer much cheaper SMARTPHONES (as opposed to the simple phones that dominate the world market)? Not likely in the near future.
Apple can never rule the world. Only companies that target things for the cheapsters can rule the world. 90% of the world consumers are poor chumps barely getting by. Apple is not targeting those people. Those people will be won over by Android (Google). Apple will make lots of money, but will never grab majority market share in anything. The analysts say iPod days are supposedly over and the music player market is dead. Google seems poised to go to $1000 a share while Apple will likely be stuck below $250 as they run out of people willing to spend extra money on Apple products.
CNN has an agenda with Apple. All of CNN's stories are positive and all are also anti-Nokia. CNN and Fortune are also linked, so this story is not surprising. CNN is to Apple what FOX is to Republican. Unbiased, right?
ex ped: Maybe you should read a little deeper in the archives, Jay.
The smartphone war ended in 2007. It just took a while for the rest of the tech world to get over its own hype and warped view of consumer interests.





This data has very little to do with market share or anything else, for that matter.
All it demonstrates is that iPhone and Android users surf the internet on their phones more than others, and are therefore more likely to prompt an ad. It's more of a demonstration of which smartphones are used for personal entertainment.
Users who source streaming media, enterprise email etc still use plenty of data, but won't show up on these graphs.