As we have some difference in our own analytics and Gamesparks I would like to understand the Retention % by day   better. 

As I see it, it could be one of 3 variants:

• The first day retention is 29 days ago, meaning if I'm interested if any changes I made had an impact on the first day retention I will have to wait 29 days to find out the result. 
• The first day retention is an average of all the "first day cohorts" that has happened in 30 days. Still meaning I must wait 30 days to get a good view of what impact any change did.
• The first day retention is always taken from the cohort one day ago, and the 30 day is based on the day 30 days ago.