Investigate late May retention gain
Categories
(Data Science :: Investigation, task)
Tracking
(Not tracked)
People
(Reporter: jmccrosky, Assigned: jmccrosky)
References
Details
Brief description of the request: Investigate early May retention gain.
Link to any assets:
Is there a specific data scientist you would like or someone who has helped to triage this request: Jesse will work on this.
|
||
Updated•6 years ago
|
|
|
Assignee | |
Comment 2•6 years ago
|
||
It appears that about half of the retention increase was attributable to the profile-per-install change and about half due to a decrease in the rate of single-day-profiles (profiles that only ever submit pings on a single day). I’m not sure why we’re seeing fewer single-day profiles. It could be that the some characteristic of 67 (performance improvements maybe?) is convincing people to keep using Firefox, but I’m skeptical that subtle improvements could drive such a significant retention increase. It might also be some artifact of how we’re controlling for the new profile-per-install profiles. I have a brief writeup here: https://docs.google.com/document/d/12S-43KquWA4WGkbhzL-y0AKySjDAGqIbU2MCW4oGfM4/
Closing this bug for now as I don't plan any further analysis of this.
Description
•