A recurring theme in many feasibility studies I do is what devices to support. This depends on factors such as what geographic region(s) you wish to distribute, the technical requirements of your application and the currently popular devices.
The latter, the current device market, can be difficult to quantify. Here are some free sources I tend to use that I have mentioned in previous posts…
- Mobile Market Research
- Statistics from companies such as Symbian
- Admob statistics
- Handango HandMark
- MobClix for iPhone
I came across a new information source last week. GetJar distribute over 33,000 separate applications and see 5,297,136 downloads per week. GetJar has downloads for Java, Symbian, Windows Mobile, BlackBerry, Palm, iPhone and Flash Lite. Hence, they have a great view of the devices adding applications. Statistics are available at mobref.com.
When I last looked, the statistics had images and data on 1302 devices. These are ranked by how new they are and their market share. Looking at the statistics, it’s interesting to see the top 19 phones are all Nokia…
If you are considering mobile development, you might conclude that Nokia S60 and S40 should be the first devices you might support. Another observation is that there’s a large long tail of phones. i.e. A large number of phones with less than 1% market share.
The ‘newest devices’ listing is particularly useful for developers already shipping applications. It gives an early indication of what new devices you might need to support.
One surprising statistic is that the most popular phone is the ancient Nokia N70-1 running S60 v2.8. This shows the latest phones may be sexy but not necessarily those adding most software. This also echo’s my experience of companies still wanting pre S60 3rd development. Older phones are cheaper and gain more market share with time, especially in developing countries (By the way, the next 4 most popular phones are actually S60 3rd).
I’d advise not looking at one set of statistics in isolation. Take a look at the other information sources at the top of this post. Each set of statistics has some skew based on the position of the particular information provider in its own market. For example, GetJar doesn’t include many iPhone applications and no Android applications both of which tend to be downloaded direct via an on-device application.