In this new age of mobile computing, the long-term success of Apple and Google depends largely on their ability to amass third-party developer support. Developer innovation improves the way consumers connect with others, entertain themselves, work, and more, all through apps. The more a platform provider can attract unique and superior content, the more appealing the hardware device appears to consumers prior to purchase and the more loyal they become afterwards.
Last week, Apple reported that it had sold a cumulative 200 million iOS devices. Currently the App Store contains more than 425,000 apps, with total downloads surpassing 15 billion. From the developer’s point of view, the most attractive aspect of the iOS consumer audience is that they all have credit cards on file with iTunes. This means 100% of them can seamlessly pay for apps and in-app purchases. All told, the App Store offers a powerful business opportunity to developers and has attracted leading mobile developer support.
At the same time, Google’s more open Android OS distribution strategy has garnered the support of numerous notable OEMs, spawning a rapidly growing installed base of Android devices that is gunning to overtake the iOS installed base. With broader distribution across more carriers, Android device activations surpassed 500,000 per day tweeted Andy Rubin last month. This growth is up from 300,000 activations per day reported just last December. In terms of apps, the Android Market has 200,000, and Google said it crossed the 4.5 billion downloaded application mark in May.
At Flurry, we regularly track developer support across the various platforms that compete for their allegiance. When companies create new projects in Flurry Analytics, they download platform-specific SDKs for their apps. Since resources are limited, the choices developers make in building for different platforms strongly signal their confidence in those platforms. They are literally investing their R&D budgets in the hopes of generating future revenue. In total, over 45,000 companies use Flurry Analytics across more than 90,000 applications. For this report, we compare Q1 to Q2 new project starts.
Studying the numbers, it’s readily apparent that Android has lost developer support to iOS. Specifically, Android new project starts have dropped from 36% in Q1 to 28% in Q2. Overall, total Flurry iOS and Android new project starts grew from 9,100 in Q1 to 10,200 in Q2. Of note, this drop in Android developer support represents the second quarter-over-quarter slide, which follows a year of significant, steady growth for the Google-built OS. Over the course of 2010, Android developer support had steadily climbed each quarter, peaking at 39% in Q4 2010.
Considering the events that could have precipitated this shift in developer support, Flurry has identified two probable causes:
1. iPhone Launch on Verizon: With iPhone’s arrival on Verizon in February 2011, three and half years after launching on AT&T, Apple closed the most significant vulnerability gap in its U.S. distribution, and likely worldwide. In fact, with its lengthy exclusive distribution agreement of iPhone on AT&T, it could be argued that Apple itself gave Android the opportunity to reach critical mass on other carriers, most notably Verizon. In that time, Google, Verizon and a host of OEMs worked hard and fast to push Android devices as an alternative to AT&T’s iPhone juggernaut. With Verizon’s launch of the iPhone, the pendulum appears to have swung back in favor of iPhone over Android development.
2. iPad 2 Launch: Establishing an installed base of more than 20 million tablet devices in less than one year, the iPad success story has been compared to taking a buzz-saw to the PC industry. Apple’s iPad shipments, from its last disclosed quarter, were higher than the initial first two quarters of iPad availability. Apple has additionally claimed that it is seeing the “mother of all backlogs.” Building efforts lag behind consumer demand for the device. We believe that wholesale consumer acceptance and adoption of tablets, which just a year ago was questionable within the industry, is further luring developers to build for iPad instead of Android.
While Android’s device installed base continues to surge, ongoing work to improve the Android Market layout and to push forward the adoption of Google Checkout are critical to its success. PayPal’s recent acquisition of mobile payment player, Zong, demonstrates that Google may not be enabling consumer payment quickly or well enough, which is inviting 3rd party competition and creating billing fragmentation. Furthermore, the development community is concerned about the rising cost of deploying across the Android installed base, due to the double whammy of OS and storefront fragmentation. With developers pinched on both sides of the revenue and cost equation, Google must tack aggressively at this stage of the race to ensure that Apple doesn’t continue to take its developer-support wind.
Although the Internet entered the mainstream a mere 15 years ago, life without it today is nearly incomprehensible. And our use of the web has rapidly changed as well. In simple terms, it has evolved from online directories (Yahoo!) to search engines (Google) and now to social media (Facebook). Built on the desktop and notebook PC platform, the web’s popularity is significant.
Today, however, a new platform shift is taking place. In 2011, for the first time, smartphone and tablet shipments exceed those of desktop and notebook shipments (source: Mary Meeker, KPCB, see slide 7). This move means a new generation of consumers expects their smartphones and tablets to come with instant broadband connectively so they, too, can connect to the Internet.
In this report, Flurry compares how daily interactive consumption has changed over the last 12 months between the web (both desktop and mobile web) and mobile native apps. For Internet consumption, we built a model using publicly available data from comScore and Alexa. For mobile application usage, we used Flurry Analytics data, now exceeding 500 million aggregated, anonymous use sessions per day across more than 85,000 applications. We estimate this accounts for approximately one third of all mobile application activity, which we scaled-up accordingly for this analysis.
Our analysis shows that, for the first time ever, daily time spent in mobile apps surpasses desktop and mobile web consumption. This stat is even more remarkable if you consider that it took less than three years for native mobile apps to achieve this level of usage, driven primarily by the popularity of iOS and Android platforms. Let’s take a look at the numbers.
The preceding chart compares the average number of minutes consumers spend per day in mobile native apps vs. the web. For mobile apps, Flurry tracks iOS, Android, BlackBerry, Windows Phone and J2ME. And for the web, our figures include the open web, Facebook and the mobile web.
Flurry found that the average user now spends 9% more time using mobile apps than the Internet. This was not the case just 12 months ago. Last year, the average user spent just under 43 minutes a day using mobile applications versus an average 64 minutes using the Internet. Growing at 91% over the last year, users now spend over 81 minutes on mobile applications per day. This growth has come primarily from more sessions per user, per day rather than a large growth in average session lengths. Time spent on the Internet has grown at a much slower rate, 16% over the last year, with users now spending 74 minutes on the Internet a day.
As a note of interest, Facebook has increasingly taken its share of time spent on the Internet, now making up 14 of the 74 minutes spent per day by consumers, or about one sixth of all Internet minutes. Considering Facebook’s recent leak regarding Project Spartan, an effort to run apps within its service on top of the mobile Safari browser, thus disintermediating Apple, it appears Facebook seeks to counter both Apple and Google’s increasing control over consumers as mobile app usage proliferates.
Games & Social Networking Dominate Mobile App Usage
With mobile app usage soaring, Flurry additionally studied which categories most occupy consumers’ time. For this snapshot, Flurry captured time spent per category from May 2011 across all apps it tracks, now totaling more than 85,000. The results are shown in the pie chart below.
The chart clearly shows that Games and Social Networking categories capture the significant majority of consumers’ time. Consumers spend nearly half their time using Games, and a third in Social Networking apps. Combined, these two categories control a whopping 79% of consumers’ total app time. Further, as we drill down into the data, consumers use these two categories more frequently, and for longer average session lengths, compared to other categories. Any way we slice it, Games and Social Networking apps deliver the most engaging experience on mobile today.
With a better understanding of how consumers spend their time across app categories, Facebook’s Project Spartan makes even more sense. As a category, social networking – which is Facebook’s core competency – commands the second largest allocation of consumers’ time. Games, which typify the most popular kind of app played on the Facebook platform itself, are also the top categories on both Android and iOS platforms. As interactive media usage continues to shift from the web to mobile apps, one thing is certain: Facebook, Apple and Google will all expend significant resources to ensure that no one company dominates owning the direct relationship with the consumer.
In 1985, Microsoft introduced the Windows operating system for personal computers. Leveraging a more open partnership model, Windows overtook the operating sytem market, with market share now exceeding 90%. As the 80s drew to a close, IBM had faltered as the world’s most valued brand in computing, leaving an opening for the combination of Windows and Intel, aka “Wintel,” to be associated with value to consumers.
In October 2008, T-Mobile released the G1 by HTC, the first commercially available mobile device running on the Android operating system. For consumers, this marked the first opportunity to use a new kind of handset born of the Open Handset Alliance, a Google-led collaboration of now more than 79 member companies who pledged to support open standards for mobile devices.
With OEMs such as HTC, Motorola, Samsung and LG all adopting the Android OS, the race is on to earn a Wintel-like place in the hearts and minds of consumers. In this report, Flurry reviews the first two years of the Android device market and concludes that the combination of Samsung and Android, or “Samdroid” as we are coining it, has emerged as a new leader in non-iOS smartphones.
Android Growth Explodes in 2010
Through its analytics service, Flurry estimates it detects over 80% of all Android devices. Presenting our own count of Android devices from 2009 to 2010, without any adjustment, the chart below shows how the launch of wave after wave of increasingly more powerful Android devices has resulted in unprecedented year-over-year growth. By our count, devices running on Android OS now exceed 60 million. From 2009 to 2010, Android adoption increased by nearly 10 times, from 5.9 million to 53 million devices.
The New 2010 Android Landscape
In 2009, the leading Android handsets were the T-Mobile G1 by HTC, myTouch 3G by HTC and Motorola Droid. Overall, for 2009, HTC’s early commitment to Android was rewarded. By the end of 2010, however, three companies split the majority of share for Android devices, with Samsung coming on very strong in the second half of 2010. Additionally, LG captured 7% of Android device activations with a surge of Q4 sales. As a new brand in consumer electronics, HTC appeared unable to withstand erosion of its early, dominant market share lead against mature brands like Motorola, Samsung and LG.
Samsung Disruption in Q4 2010
Breaking out new Android device activations by OEM by quarter in 2010, as is done in the subsequent chart, provides a view into how Samsung enters strongly in the back half of 2010. In fact, Q4 2010 marked the first quarter since Android launched eight quarters ago in Q4 2008, with the T-Mobile G1 by HTC on T-Mobile, that HTC did not lead the market in quarterly new Android activations. While HTC enjoyed strong first-mover advantage, coupled with additional penetration through its willingness to white label for Google and T-Mobile branded Android devices, it succumbed to the brand power of Samsung led by its Galaxy S line of Android devices in Q4 2010.
The Galaxy is out of this World for Holiday 2010
Studying the Holiday period in December 2010, not only does Samsung dominate the top position with the Galaxy S, but also claims the third best spot with the Galaxy Tab, the best-selling Android tablet and only non-phone in the top ten. Additionally noteworthy is that the top five does not include HTC. However, it’s also fair to point out that by expanding the list to include the top ten devices, spots six through ten are all HTC devices, in the following order: HTC EVO 4G, HTC Desire, HTC Incredible, HTC Glacier and HTC Wildfire.
Flurry Estimates 20,000 First Week Nexus One Sales:
Not an Apples to "Apple" Comparison
The Google Nexus One launch has become the most controversial and confusing Android handset launch to date. With publicity "leaks" over the holiday season, the Nexus One handset received unprecedented buzz. This same hype helped create the expectation of a revolutionary Android handset and its potential to be an iPhone killer. To gauge the success of Google's first handset launch, Flurry leveraged its analytics reach to estimate launch week sales of the Nexus One.
Flurry monitors usage of more than 10,000 developers' applications on iPhone and Android platforms. In total, Flurry tracks applications on approximately four out of every five iPhone and Android handsets in the market, generating over 25 million end user sessions per day. To estimate first week sales totals for the Nexus One, myTouch 3G, Droid and iPhone 3GS, Flurry detected new handsets within its system, and then made adjustments to account for varying levels of Flurry application penetration by handset. Flurry additionally crosschecked its estimates against Apple actual sales, released for iPhone 3GS, which totaled more than one million units over the three days, June 19 - 21, 2009. Flurry first week sales estimates can be found in the table below.
While Flurry estimates that Nexus One was outsold by Droid by more than 12 times, myTouch 3G by 3 times and iPhone 3GS by a staggering 80 times, it's worth noting there are significant differences in the marketing, distribution and perception of the device as revolutionary vs. evolutionary. In short, key business decisions and other factors related to the Nexus One launch make an "apples-to-apples" comparison difficult.
As a product, the Nexus One boasts the most advanced Android OS to date as well as unique features, such as Google Voice and Google Maps. However, potentially due to the heightened "promise" created by early buzz, the handset has ultimately fallen short on sales expectations. Without the "wow factor" now expected with each new challenger to the iPhone, especially the first smartphone with Google's own branding, demand generation has been modest.
Next there is distribution, pricing and marketing to consider. For its release, Google executed an online "soft launch" of the Nexus One, a very different go-to-market strategy compared to Verizon's launch of Droid, on which it spent a record-breaking $100 million on marketing, including aggressive TV advertising spends. Instead, Google chose to market and sell the device to consumers directly through its own website. While this distribution strategy is among the most innovative facets of the Nexus One launch, and a threat to carrier control of the consumer relationship, a series of customer service and other mistakes reveal Google's lack of retail experience. Further, Google chose to launch its Nexus One after the holiday season. While selecting this launch time may have been designed to avoid competing unnecessarily with its carrier and OEM partners over the all-important holiday season, this also adversely affected sales.
Additionally, T-Mobile, Google's carrier partner for Nexus One, did not provide the same carrier co-marketing support as it did for the myTouch 3G launch. Cannibalization may also be playing a role as the Nexus One competes against the myTouch 3G for any new T-Mobile customer. In effect, sales are now split between the two handsets. And while Google, in an effort to avoid channel conflict with T-Mobile, appears to have set the direct-to-consumer price for the handset at over $500 dollars, the high price point combined with the fact that the handset is only considered an "evolutionary" improvement over previous Android devices, indicates that Google did not take the steps to maximize first week sales. This is especially evident when one considers that among the most expensive costs associated with the launch - marketing - has not been incurred, and could have been applied to lowering the direct-to-consumer price point.
With each consecutive Android launch, consumers are enjoying more choices in the market. The flipside of this, however, is that it will take increasing innovation and decreasing price points to attract new smartphone consumers. Further, for Android to continue to chip away at Apple's third-party app developer support - key to delivering consumers more value - it needs to aggressively grow the Android device installed base. Only then will application developers more fully support Android.
There is a lot riding on the Motorola Droid. Verizon is looking for an answer to the iPhone, which has driven enviable data ARPU growth for AT&T. According to Rita Chang of Ad Age, the Droid is supported by a $100 million integrated marketing campaign, the largest in Verizon history, running through the end of 2009. Google's long-term bet on mobile, underscored by its recent $750 million offer to purchase Admob, demands that the Android OS propagate. Meanwhile, HTC enjoys an early mover advantage with its G1 and 1.5 generation of Android handsets. Finally, several more leading OEM's have pledged to support Android, with as many as 100 handsets scheduled to ship in 2010. In short, the once dominant, now reinvented Motorola has little room for error.
Through its analytics service, Flurry monitors usage of over 10,000 developers' applications on iPhone and Android. In total, Flurry tracks applications on approximately two out of every three unique iPhone and Android handsets in the market, including over 15,000 million user sessions per day. To estimate first week sales totals for the myTouch 3G, Droid and iPhone 3GS, Flurry detected new handsets within its system, and then made adjustments to account for varying levels of Flurry application penetration by handset. Flurry additionally cross-checked its estimates against Apple actual sales, released for the iPhone 3GS, which totaled one million units sold over the three days of sales, Jun 19 - 21. Flurry first week sales estimates can be found in the table below.
While Flurry estimates Apple sold approximately 1.6 million 3GS units over its first week of sales, it is important to note that Apple simultaneously launched its device across eight countries (U.S., Canada and six European countries), while the Droid launched only in the U.S. Additionally, the iPhone commanded a strong installed base of over 25 million at the time the 3GS launched. Of those, over 6 million were first generation iPhone users who were expected to upgrade to the 3GS. Taken in this context, Droid sales of 250,000 units during its first week from a standing start and in just one country, is a strong result for Motorola and Verizon. Also, by Flurry's measurement Android does have an edge over iPhone app usage, with the average Android session length at four minutes vs. two minutes for iPhone apps.
With its first week showing, the Droid is the fastest-selling Android phone to date. Compared to the myTouch 3G on T-Mobile, Droid outsold it by more than four times. Of course, Verizon has nearly 90 million subscribers to T-Mobile's 34 million and Verizon has shown a willingness to outspend previous carriers with its $100 million campaign. With Motorola forecasting its own Droid sales at one million through the end of the year, that equates to a cost per acquisition of $100 per handset sold. Verizon's aggressive campaign coupled with its 3G network advantage in key, populous regions of the country like the Northeast positions the Droid as a legitimate challenger to the iPhone.
While iPhone continues to dominate the Smartphone market overall, each subsequent Android handset launch increases competition for Apple. Furthermore, Flurry has monitored a sharp increase in new Android project starts by application developers within its system, a 94% jump in October compared to September. The launch of Droid signals the beginning of a viable platform alternative to the iPhone as Android builds critical mass. As major companies continue to vie for a piece of the exploding Smartphone market, the consumer has never had more choice and innovation in the mobile industry. With Droid, Motorola has raised the bar for Android handsets, contributing to an ever-growing base of Android handsets upon which applications developers can build a business.