Thanks to a link on daring fireball, I read this entertaining comment thread on the BBC’s website. Mostly, it consists of Android users complaining that the BBC releases apps on iOS first. But the BBC comes back with a very even-handed and articulate defense of their releases (emphasis added is mine):
Thanks for your comments so far.
On the Apple v Android question. It is absolutely our plan to have quality products out on both products as soon as possible. There are three main factors we had to keep in mind when lining up releases: development, testing and launch complexity.
As the BBC Sport app is a hybrid app, based on the new Sport mobile browser site, the platform-specific development can progress in parallel, building on the core browser site. The decision to launch the core mobile browser site first (before either app) was itself to ensure that users got a quality product across as wide a range of devices as possible. The Android-specific development is very close to completion.
Due to the huge range of Android devices, testing for that platform is more complex and therefore takes more time.
And finally, scheduling ‘big bang’ launches, across a range of platforms, increases risk, and we want to ensure launches are as smooth as possible for users. Back in July, when we launched the Olympics app for iPhone and Android together, we saw over three times as many downloads of the iPhone version. Android continues to grow apace but this, together with the development and testing complexity, led us to the decision to phase the iOS app first.
Features also add complexity (and risk) to launches, which is why we’ll be adding football teams, team customisation and video in the coming weeks, across all platforms, rather than at launch.
So. To sum up – developing for Android devices takes longer, requires more testing, and appears to be more expensive. And yet, more BBC customers download the iPhone app than the Android app… By a factor of three! And finally, rolling out to a homogenous platform first allows for a smaller overall roll-out – minimizing the number of risk factors, as well as somewhat limiting the audience.
The comment stream does a good job explaining why major enterprises, not just startups, are still choosing to develop on iOS first.
From a process point of view, iOS first fulfills a few directives:
- Simplify the deployment scenarios (environment risk)
- Reduce non-functional scope (supporting myriad devices is extra scope)
- Minimize testing overhead
- Focus on requirements first, then worry about platform support
- Avoid big-bang roll-outs