Many clients mistakenly believe that supporting previous versions of iOS software is important. But almost universally, supporting iOS 4 and earlier is a bad investment.
Most iOS users can upgrade
Unlike other platforms where software updates are provided for very limited windows, Apple’s updates run on devices up to two years old.
Apple works hard to make new versions of iOS available to an unprecedented number of users, even those users who may be running older devices. The vast majority of your potential customers don’t need new hardware to run iOS 5.
Most iOS users have already upgraded
Within 5 days of launch, iOS 5 was in use by 33% of devices. Marco reported in November that 48% of his customers were running iOS 5. Bump reports that iOS 5 penetration is about 60%, (with the next-oldest OS, 4.3.3, being a paltry 9%). Our internal statistics show that our customers are close to 70%-80% iOS 5 users.
And don’t underestimate the length of your development cycle. New applications often take 90+ days to develop, and by the time your application actually ships, the numbers will be even higher.
Supporting iOS 4 is expensive
Supporting iOS 4 means double the test burden, means that everyone has to buy extra devices to test older configurations, and means that the cost to test your software triples. These are the obvious costs.
Supporting iOS 4 has hidden costs too: it means that great developer features introduced in the last 6 months can’t be used in your application. This increases development time and cost significantly, as complex “workarounds” have to be created to avoid the use of these features. These design choices can have long-lasting permanent effects on your software’s code base that will far outlive iOS 4.
But there’s another hidden cost: supporting iOS 4 means using fewer iOS 5-specific features. Apple users expect a high-value experience on their shiny new Apple device, and if your application fails to deliver the latest and greatest, you’ll get bad reviews. If you have a competitor who produces a more highly-featured application, expect a large portion of your customer base to pass over your app in favor of the competing one with iCloud support and Notification Center integration, Twitter integration, and other iOS 5-only capabilities.
Supporting iOS 4 is bad for business
At this point, most of the iOS 4 holdouts are either people with very old devices (cheapskates unlikely to buy your product), or users who are technically incapable of performing a software update (and who will present a high technical support burden). These are not the kind of customers you want to be burdened with.
Supporting iOS 4 can be attractive in a requirements list, but it can be a costly exercise that provides no real revenue to justify its cost. Many developers make good money supporting iOS 4 at a markup, but in the best interests of our clients, we recommend that the vast majority of new applications be written to support a minimum of iOS 5, and that significant (non-bugfix) updates to existing apps seriously consider dropping support for older versions of the OS. Focusing on iOS 5 provides the best value for most projects, and redirecting legacy costs into new feature development can be much more exciting for both you and your customers.