App Analytics - User Behavior analysis (per user, instead of session)[ Edited ]
April 2016 - last edited April 2016
We've been trying to implement Analytics in our mobile game, but there are some information we would like to obtain about user behaviour (And it looks like Analytics could do it) but we just don't know how.
First: Flows per user, not per session
Is there a way to see the flow per user (in his lifetime) and not during each sessions?
The game is an exploration game, where each "level or scene" of the game is a screen for analytics. It's a good info to understand player's flow in a session. But it would be much more interesting for us to know user's flow in his life using the app.
This way we would know where players are stopping to play the game "for ever", and not only when they stop for a nap or anything else. We would also be able to understand the paths that are usually taken by the users while exploring the game, since the player can stop the game for a pause and reload it in a much more advanced screen. (this creates many entry points per session, and it is much harder to understand the path the player took to get in there).
Second: Events flow (filtering per action, or category)
Is there a way to check the flow only for events in a Category or in an Action?
The way analytics shows now, it messes up in game Story Events with Button Clicks Events with Player Death Events... which are quite unrelated, and it is even worse because they are show per sessions and not per user..
Third: When defining Goals and funnels, can't I use Events as a destination Goal, or to check for the steps?
I'm not sure if any of this is possible, I understand Analytics is made with websites and "websites like" apps in mind. But it looks like it has all the info and processing power we need.. we're just missing some filtering so we can visualize this info the way we need to.
Re: App Analytics - User Behavior analysis (per user, instead of session)
You will want to review Triggers and Event Objects: https://developers.google.com/apps-script/guides/triggers/
Depending on the signup/sign-in process for the App, you might be able to use Client ID in Analytics. https://goo.gl/2DOaV2
Ultimately though, you might choose to separate the usage tracking by keeping user data and checkpoints in a separate database. This is the BigData approach. You could send the Analytics data, the event and install data, other information to Google Big Query API to better track Event flow and usage information. This would be a large undertaking, but would make correlating information in the future. https://cloud.google.com/bigquery/what-is-bigquery
If you don't want to remake the wheel, I would look at Mobile App Development platforms to see what kind of additional tracking and analytics they offer. This is not an endorsement, but Steam Powered is a popular platform: http://www.steampowered.com/steamworks/developmenttools.php