We give our app users 100 points for signing up, 20 points every day just for checking in, 30 points for taking a survey, and 400 points for playing a game. 

When users get to 500 points, they can cash out for prizes. . 

Question: How can I find out the average breakage by feature? 

For example, if the retention rate on D1 is 50% and users who churn never reach 500 points, then most welcome points are “broken” and are never used to cash out for prizes, so the breakage on welcome points is very high, maybe 90%. Meanwhile, most users who play games are earning enough for prizes, so the breakage on games is very low, maybe only 10%. 

Any user that is dormant for 90 days we can consider lost and all their points broken.  

We have all the data in Amplitude (# points earned by user for what activity, # points broken), but since all points all go into the same points balance, I wonder if there’s a way to chart and estimate of breakage by feature type in Amplitude. 

Thanks for your tips!