According to the latest Nielsen numbers, Android, iOS, and BlackBerry are nearly neck-and-neck in smartphone OS market share with 29%, 27%, and 27% respectively. When looking at the graph, one thing stands out in comparing new leader Android to its nearest neighbors…Apple and RIM are putting all their efforts in two fronts: OS and hardware. Meanwhile, we note that the ever-increasing Android footprint is actually enabled by a slew of hardware manufacturers running the ‘open’ Google operating system.

You hear ‘open’ a lot when Android is discussed. The word is probably not on the mind of your average non-technical consumer when making a phone purchase, but it is definitely a key differentiator with the tech types who would prefer Android over another. After all, Android is based on an ‘open’ Linux OS with an ‘open’ hardware standard, open to all manufacturers who usually provide such ‘open’ features as removable storage and a replaceable battery.

Apple – and to a lesser degree, RIM – is the target of this open attack. After all, the iOS ecosystem in particular is a ‘closed’ one – devices with “no user-serviceable parts inside” supporting a “walled garden” that ensures a consistent user experience by preventing developers from straying from established standards and closely vetting apps to ensure uniform behavior.

The other word you hear a lot is ‘fragmentation’, thrown out by those who believe Android’s ‘open’ advantages also have potential disadvantages. But what exactly does fragmentation mean? If you allow examination with a very liberal definition, I think it can mean far more than originally intended.

OS Fragmentation

The original concern over fragmentation arose because of the varying versions of operating systems that were shipping on Android phones, inconsistently applied. A developer writing an app for the predominant v2.2 would have to be prepared for leading-edge Gingerbread (2.3) and legacy v2.1 or earlier. Or viewed from the other side, a very current Gingerbread-era app has to contend with backwards compatibility with the previous versions in order to work on the majority of devices.

The problem with diversity of versions is magnified by two critical factors: some devices are forever locked on certain releases because of an inability to upgrade, and the rest are on upgrade schedules determined by the carriers (rather than Google or the handset manufacturers). Carriers are not obligated to deploy a new OS, even long after Google has released it. Perhaps this explains why, as of Jan 2011, only 0.4% of Android devices were on the “latest” version and 52% were one rev behind. So nearly half are on “very old” releases.

Apple is certainly no stranger to different OS versions., but the state of iOS adoption for the current+previous is much stronger than with Android, at >80%. Because Apple controls the release schedule and deploys centrally via iTunes, more people comply. And because Apple limits the number of platforms and the impact of variability within, compatibility complications are minimized. In this case, the “walled garden” has some advantages.

Hardware/Carrier Fragmentation

It seems like there is a new Android phone out every month. The cadre of Android manufacturers puts Apple somewhat at a competitive disadvantage because they can’t immediately integrate the latest features that the market dictates. Unable to churn out hardware as quickly as the collective efforts of HTC, Samsung, Motorola, and LG, iPhone users have to be content with an annual hardware release.

But this constant flux is what complicates the Android OS fragmentation: the hardware itself is fragmented. Capacitive buttons (or not), video camera (or not), HDMI out (or not) – these differentiators have the potential to change the user experience. AMOLED screens, NFC sensors, and other manufacturer choices require drivers to support their operation, and this drives the churn in the OS which leaves previous releases in the dust. Developers must code to this moving target as some devices go forward and others stay stagnant. And when carriers then add their own skins and applications, it often means that an Android user can’t simply pick up a different Android phone and use everything without a brief period of familiarization.

It’s not to say that Apple does not have to contend with the same type of platform evolution, but even the GSM iPhone2 experience is nearly identical to the CDMA iPhone4. Because the hardware is controlled by a single manufacturer who also provides the development environment, it allows apps to maintain compatibility in both directions. Indeed, the number of iPhone apps that are limited by OS version are in the extreme minority – those few whose existence depends on a newer hardware feature or OS behavior. (Of course, I’m sure it also doesn’t hurt that developers know they would only alienate a small fraction – a mere 0.02% are more than one major iOS rev back – when deciding to make their app be version-sensitive.)

In no place can this be better exemplified than with the 2010 release of the iPad tablet. Day One users of the very first iPad had full access to all iPhone apps – they ran in an iPhone-sized portion of the screen and did not take advantage of the iPad features, but they ran nonetheless. Contrast this with the release of the Motorola Xoom with the tablet-specific Honeycomb version of the Android OS. At the time of release, there were few Honeycomb’d apps that would run on the device, and many regular smartphone apps would crash when run on a tablet footprint – new hardware, new OS.

Google’s evolution has taken Android down a split path: the large-screen Honeycomb is the optimized tablet OS, and the small-screen Gingerbread is for smartphones. Users will have to wait for “Ice Cream” before they can have a universal OS across both platforms. When you add a new dimension of “device type” to the equation, the issue of “fragmentation” becomes even greater than if just for smartphones.

Now as would be expected, a docking station for a Dell Streak 5 is not compatible with a Samsung Galaxy Tab. One would not presume a level of interoperability across competitor Android devices. Yet a docking station for a Dell Streak 5 also is not compatible with a Dell Streak 7. This is certainly not the case with Apple, where the dock connector has remained relatively unchanged through the years. Not only will first-gen iPad accessories also support the newer iPad2, but in many cases the iPod and iPhone as well. Android devices may be based on ‘open’ USB, HDMI, and 2.5mm audio standards, but because their port placement is ‘open’ to interpretation, there are no universal Android accessories despite having surpassed the iPhone’s market share. The fragmentation of the hardware means that there are limited technical synergies and few incentives to develop after-market accessories.

Some cite the Apple dock connector as being ‘proprietary’, yet I would ask: which accessory ecosystem is more ‘open’ when interoperability with different device types is possible? With Apple’s ‘closed’ hardware, they can provide a consistent OS that can run on all devices that they sell, supporting a robust after-market of nearly universal 3rd-party hardware accessories.

Marketplace Fragmentation

The iTunes AppStore is much maligned, and often rightfully so. Their policies can sometimes be anti-competitive and not always in the best interest of what the customer wants. But their processes ensure a consistent user experience and there has yet to be a sighting of iOS malware (that wasn’t in a jailbroken environment). The safety and consistency afforded by the decidedly unfragmented AppStore is what most Apple users have come to rely upon.

And possibly the best thing for the consumer about the AppStore is that it is a one-stop shop for all things iOS. A single app download is valid for as many iOS devices that the user owns, which is quite a contrast to the old-school fragmented PC/Mac licensing model.

Android users do not have the luxury – though some would say ‘handicap’ – of a one-stop AppStore. Instead, there are a competing group of ‘Marketplaces’ run by Google, Amazon, and the individual carriers. These stores are ‘open’ in the sense that they have fewer restrictions on what can be sold, and they are not under the totalitarian rule of a single vendor.

But the openness brings with it inconsistencies, and the less stringent curation has occasionally allowed instances of virii and phishing exploits to slip through. Android was successful in one-upping Apple with their WiFi app transfer scheme, but the license portability is suspect and still dependent on the source Marketplace, making universal licensing difficult.

Brand Fragmentation

People are often critical of Apple for lack of choice when it comes to their product catalog. Aside from memory variations, there was initially only one iPod, one iPhone. The iPad was possibly their most diverse first-run product: with or without 3G. Multiple SKUs are simply not a common characteristic of the Apple brand on first-out products. They often design a single version and trust(hope) it meets consumer acceptance. It doesn’t always work (Newton, AppleTV). But by focusing on a single SKU on first release, one advantage is that Apple can put all their efforts into optimization of their R&D, system engineering, and supply chain, often resulting in a product that is cheaper to manufacture than their competitors.

Other manufacturers typically flood the market with multiple SKUs – great for consumer choice, albeit confusing. In the 2 year old Android handset market, the top 3 manufacturers have a combined 45 models available across all carriers. Has this product diversity fragmented the manufacturers’ engineering and supply chain efforts? Not sure. But it likely affects the marketing, given that the Android brand is now an amorphous blend of Heros, Droids, and Galaxies.

The variety of hardware for tablets in particular has a huge potential to negatively influence the Android brand. Because there is no requirement for a tablet to have 3G/4G cellular, the barrier to market entry is smaller when distribution is not tied to the carriers. Any manufacturer who thinks they can crank out a WiFi netbook can get into the tablet game. But if a $99 tablet proves to be less than stellar, what does that do to the consumer impression about “Android”? Are they likely to later spend many times that amount on a better performing Android tablet? Or are they going to let the bad experience tarnish their view of the platform and spend their money on the premium Apple brand (especially when ‘premium’ may no longer mean “more expensive”)? The open market allows familial competitors to undercut, fragmenting the Android brand.

As for their own brand, Apple is shedding their elitist label and moving into new lower-end markets. Their publicity engine is intent on telling a unified story that hardware plus software equals a superior user experience. ‘Closed’, yes, but they would have you believe that it is all for the benefit of the user. Every advertisement is about usage and experience, not “speeds and feeds” as Steve Jobs claimed in the iPad2 release party.

Spend a few hours with commercial media and you will notice that Android manufacturers and carriers as a whole spend more advertising dollars than Apple does (perhaps 10:1 or 15:1?). It is questionable whether this onslaught enhances the Android brand or not. Very few actually highlight the “doing”, and instead it becomes about why one Android device is better than another. Most are not yet operating in what Apple disingenuously declared the “post PC” era – instead, every HTC/Samsung/Moto commercial is all about more speed, more megapixels, more memory.

And while Android manufacturers are busy fragmenting their marketing efforts by competing head-to-head on specs, Apple can often just sit back and let others do their marketing for them. For instance, pick up any Brookstone or Sharper Image catalog and bear witness to a tremendous brand-boosting campaign that Apple marketing had no hand in. Such publications advertise dozens of gizmos that are expressly designed to dock with an iPod/iPhone, and in some cases, iPad – further advertising for the Apple brand. The same cannot be said for Android, despite superior numbers. If an HTC Hero or Samsung Galaxy user wants a premium speaker system, MP3-enabled clock radio, or a cycling trainer, what are they to do? The iPod/iPhone/iPad open after-market is a very compelling consumer story.

The Fallacy of Open vs Closed

‘Open’ and ‘closed’ may be traits of the platforms, but neither is inherently good or evil – they are barely even purchasing factors. Openness is certainly real, but in and of itself, it is mostly an importance to the geek crowd. (And they can’t claim ‘open’ without acknowledging ‘fragmentation’ as well.)

So what influences brand acceptance and revenue generation, if not the the level of openness? It is what that openness (or lack thereof) ultimately contributes. The consumer is attracted by performance, capability, security, user experience, variety of applications, and availability of after-market accessories. ‘Open’ and ‘closed’ can mean different things to users when viewed in those other dimensions.

“Speeds and feeds” has certainly not gone by the wayside (yet), and I’ve seen Best Buy customers swayed by megapixel arguments. But I’ve never heard one say “yes, but is it ‘open’?”.