Analyzing Flurry data

Have a mobile app that sends data to Flurry? Would you like to do some custom analysis on that data? Splunk to the rescue!

The new Flurry App for Splunk provides a scripted input that automatically extracts events from an existing Flurry account.

A few example questions you can answer:

  • What days is my app used on?
index=flurry | eval weekday=strftime(_time, "%w") | stats count, values(date_wday) as weekday_name by weekday | sort +weekday | table weekday_name, count
  • When during the day is my app used?
index=flurry | stats count by date_hour | sort +date_hour
  • Who are my most active individual users?
index=flurry Session_Index=1 | top User_ID

And you can answer custom questions that are specific to your app. For example, for one of my own apps, I can answer questions like:

  • What screens on the app are used most often?
index=flurry | top ShowView__View

  • Where are my users located?
index=flurry SearchFinishWithLocation__UserLocation=* | fields SearchFinishWithLocation__UserLocation | eval _geo=SearchFinishWithLocation__UserLocation

  • Are users converting?
index=flurry (ShowView__View="PPHomeViewController" OR ShowView__View="PPSearchListViewController" OR ShowView__View="PPSearchMapViewController" OR ShowView__View="PPSpotDetailsViewController" OR SpotListSpotsChanged__NumFilteredSpots>0) | eval ShowView__View=if(SpotListSpotsChanged__NumFilteredSpots>0, "GOT_RESULTS", ShowView__View) | fields User_ID, ShowView__View | timechart span=1w dc(User_ID) by ShowView__View | eval SEARCH=PPSearchListViewController+PPSearchMapViewController | fields _* PPHomeViewController SEARCH GOT_RESULTS PPSpotDetailsViewController | eval PPHomeViewController=PPHomeViewController-SEARCH | eval SEARCH=SEARCH-GOT_RESULTS | eval GOT_RESULTS=GOT_RESULTS-PPSpotDetailsViewController

Download the app today and let us know what you think on Splunkbase.

David Foster

Posted by