There is also a HockeyApp . I compare Flurry and HockeyApp based solely on distributing my propaganda on the site, and my resume is that HockeyApp is more "fix-oriented" while Flurry is more "sell-oriented": Dev and QA will win more from HockeyApp., and Product Management would benefit more from Flurry's insane Slice-and-Dice analytics. Hope this helps you based on what you are trying to accomplish.
Update . I had a quick chat with the guys of Crittercism and I wanted to add my conclusions. It seems like their offer filled the gap between HockeyApp fixed hockey and Flurry sales-oriented. It uses the same PLCrashReporter core library to generate reliable crash reports like HockeyApp, and it looks like they have more Flurry-like analytics than HockeyApp. Also, prices ... Flurry is free (although they seem to be monetized through ads on their WWW interface). HockeyApp has pricing based on plans starting at $ 10 / month. Price criterion based on the # active users of your application, and you must work with your sales staff to find out the actual quantity.
Also update support : HockeyApp support is excellent; I never waited more than 10 minutes to get an answer to my questions, and the answers are brief and accurate. Flurry, about 24 hours of turnover and a rather unlicensed and lifeless response that was mostly accurate. The criterion responded fairly quickly to my requests; they have a "chat now!" who handed me over to WHO, which was very similar to the technical issues that I had.
Some key features and developments:
HockeyApp automatically symbolizes crashreports of users on its network interface and can group failures upon API failure. Flurry does not; it just shows you a bunch of raw failures, and you get the option to manually symbolize them one command at a time (using atos - you cannot use symbolicatecrash because Flurry does not give you the correct. failure report). Keep in mind that line numbers in HockeyApp can be disabled on one or two lines .
A crash report reports that Flurry indicates that you are not including other running processes, whereas HockeyApp. In fact, it seems that Flurry crash reports are truncated to 255 characters or so. They are anemic compared to HockeyApp.
In the event of failures that you mark as fixed in the HockeyApp web interface, HockeyApp can inform a user who subsequently experiences such a crash that the problem has been fixed in the new version.
Flurry has a very deep analytical prowess that HockeyApp cannot touch: tracking usage statistics, customer interactions, average session duration, geographic distribution of your application, and user retention over time.
The storm is free; HockeyApp (free up to 10 apps). Both provide email support, but only HockeyApp offers a discussion group. Both can record arbitrary messages (for example, a JSON response from the server that caused your application to fail), but only HockeyApp indicates that this message can be of any length.
Alas, these are just a few random tidbits that my nature of the developer liked and made me prefer HockeyApp. I wonder what will happen if I use both in my application!
If you have 7 minutes, HockeyApp has a video walkthrough that I found quite useful.