We believe users who reset their password, especially single device users, may have a bad experience. We don't know how this affects their sync usage. Before we decide what a good metric to track is, it will be helpful to look at the data in a more raw form to see what patterns and relationships there are. From there hopefully we can identify a metric that we can strive to improve. Questions to answer: - what's the difference in activity before and after a reset? - days of usage - amount of data synced - number of active devices - are there different patterns in those numbers between single and multi-device users? - what do good and bad reset experiences look like in the data? Method: - Gather all the device_activity for 1 week prior and after a reset for a set of users. - Divide users into single and multi device groups (before reset) - Identify obvious cases of bad and good experiences - Compare data of each user from those cases - Identify criteria an bounds for tracking metric(s)
You need to log in before you can comment on or make changes to this bug.