How can Android win developers?

The first production Android device (T-Mobile's G1) will soon hit the market, and the legions of Google faithful will no-doubt snap up many devices in short order.

Will that be enough to allow Android to be a real competitor in the smartphone wars? The G1 isn't a particularly attractive device, doesn't have as good a media player as the iPhone, doesn't have a browser that can best the iPhone, doesn't provide eMail that holds a candle to BlackBerry, and is locked to a network that's widely regarded as having lackluster coverage. It's unlikely that the G1 is going to be a hit with consumers, and it's not going to ship in volumes that will make anyone at Apple or RIM lose sleep.

That feeds a bigger problem, though: getting developers on side. Unlike the iPhone which got great consumer demand without any third-party apps just by having a sexy device, or BlackBerry which has thrived by providing a rock-solid communications "appliance" that nobody's yet rivaled, Android is going to need to differentiate on killer 3rd party apps.

When it comes to developer platforms, iPhone has a beautiful SDK and distribution mechanism that's sparked widespread adoration from developers. Sure, Apple's exerting far more control on the developer community than perhaps they should, bit they're still a whole lot better than the existing carrier / OEM ecosystem. Also, iPhone is a single, consistent platform... a developer can build one app and have it reliably run on the millions of iPhone 2G, 3G and iPod Touch devices in market. A great, powerful developer toolkit, robust operating system, ubiquitous distribution platform, and consistent device hardware? This is a developer's dream.

The BlackBerry developer platform isn't particularly elegant, doesn't have a great development environment (RIM just recently got decent Eclipse integration), and is a generally underpowered OS when it comes to doing anything involving graphics or media. It's also got a pretty significant number of inconsistencies between OS and hardware versions, which has proved to be an irritation for many RIM developers. It's particularly telling that RIM wrote the Facebook app for BlackBerry, whereas Facebook built their own iPhone app.

That said, BlackBerry provides a reasonably reliable and stable operating system, and millions of devices as an addressable install base (with users who have a disproportionately high usage of data and third party apps vs all the other platforms except iPhone). This makes it a no-brainer for many application developers to target.

So a developer just developing a new application will probably build their first app on iPhone, do a re-write in Java to support BlackBerry, and then give very careful consideration to whether they feel like building an app for Symbian or Windows Mobile. In a competitive marketplace, it's safe to say that if you can hit 30-40MM devices by targeting iPhone and BlackBerry, it's unlikely that you're going to spend anywhere near the same resources to target a platform like Android with <1 MM. Those resources are probably better spent on building either for a platform like Symbian (with broad reach), or towards building version 2 of the iPhone/BlackBerry app.

What's Google to do? Well, how about building a porting toolkit? If a developer decides to build a BlackBerry application, they're already going to be creating a Java version of their application. If Google provided tools to help more quickly port the BlackBerry Java app to an Android Java App, that could significantly reduce the time and effort required to get an app on Android, making the business case for doing so much more viable.

Sure, some classes of apps that are really glued to the hardware (e.g. games), aren't good candidates for something like this, but those apps are in the minority. Also, while Android's Java implementation has all sorts of great features like 3D graphics, rich sound libraries, and more, BlackBerry doesn't, so any app that was a straight port from BlackBerry to Android wouldn't really show off the full potential of the Android platform. Not ideal. But if the porting toolkit were extended to quickly and easily add some of these richer capabilities into the resulting applications, it could prove a powerful motivator for developers.

The name of the game is addressable market, and it's clear that as the new kid in town, Google's got an uphill battle. With a far-from-inspiring first device, the battle's just gotten a bit harder. Taking advantage of the current strength of BlackBerry as a platform is one thing that Google can do to more quickly bootstrap their own developer ecosystem... which would be great for the entire industry.