For those that are interested, here is recent (well actually, today’s) CSV file with the raw data that I’ve been using to look more closely at the App Store. Feel free to use it for good.
If you do decide to blog about something interesting that you’ve found in analyzing the data, it would be great if you could give me a little shout out/link and also if you could let me know so I can benefit from what you’ve learned!
How is popularity calculated?
Posted by: Brandon Turner | December 19, 2008 at 05:19 PM
It's a big secret- that is a value that Apple calculates and uses to represent popularity in iTunes. They haven't revealed where that comes from, though I can tell you that it ranges from 0 to 1 and it appears that there is a 'most popular' application in each of the categories, which scores a 1. This means that comparisons across categories aren't interesting since the popularity value is relative to the category that the application is in...
Congrats on Writer- it is awesome.
-c
Posted by: Charles Teague | December 19, 2008 at 06:06 PM
Thanks.
I wrote a small post which expanded on something you talked about briefly in an earlier post where you said free applications are ruling the app store.
http://blogs.msdn.com/brandonturner/archive/2008/12/19/it-might-cost-you-to-be-popular.aspx
Posted by: Brandon Turner | December 19, 2008 at 08:37 PM
Charles,
Great analysis. I'd like to reference your work in an upcoming research study. Please contact me about obtaining the latest CSV.
A couple thoughts on trends:
1) there seems to be a fair amount of "gaming" by developers to improve rankings on the App Store. Developers and publishers I spoke with see a lot of single-feature apps emerging instead of feature-rich apps. The recent open letter to Steve Jobs from a developer reiterates this concern.
2) most of the apps available today are in English. An opportunity for homolgated versions exists. There is a business to this -- a lot more thought into support and community is needed in addition to translating text strings. The side-effect would be a spike in total app count -- similar concerns about counting "real" apps exists with multiple releases, trial vs. paid, and now localization.
Posted by: Jeff Orr | December 31, 2008 at 02:34 PM