Solved

What is the difference between 30-day retention and 1 month retention?

  • 3 September 2020
  • 2 replies
  • 52 views

Badge

1. What is the difference in the way unbounded retention is calculated for 30 day retention vs 1 month retention?

2. What is a use case for insights that you would want from 1-month, 2-month etc unbounded retention (instead of using 30-day, 60 day unbounded retention etc)?

 

E.g. scenario

 

Average 30-day retention (unbounded) for this cohort of users in May is 15%.

Whereas 1 month retention (unbounded) for this same cohort of users in May is 35%.

 

My understanding is that 30-day retention (unbounded) is looking at users who have repeat use on the 30th day or later. And then this result is averaged for the cohort in the selected time period (in this case May 2020). 

 

But shouldn’t 1 month unbounded retention also occur in the time bucket of after each user hits 30 days (?or 28 days/or 31 days) or anytime thereafter i.e. shouldn’t it the result be the same/similar to 30-day unbounded retention?

 

Unless is 1-month retention looking at next calendar month rather than 1 month (30 days) since event A? (e.g. user did event A on 28th April and then returned on 2nd of May to do event B - thus would fall into ‘next calendar month’ bucket?

 

The use-case for 30-day, 60-day, 90-day etc unbounded retention I understand - how long are engaged users sticking around for and are continuing to be engaged.

What is the use-case of 1-month, 2 month unbounded retention?

icon

Best answer by Jaya 8 September 2020, 14:16

For your second question:

30-day unbounded retention means activity on 30th day or later.

1-month unbounded retention means activity in first month or later.

In the second case, everyone with activity in the entire first month is also counted. In the first case, the first 29 days won’t be counted.

For your second question:

The report does not show the prediction, but what is the current status. 9.33% of people who returned on 29th days, have not yet returned after that. If you are not too far away from that date, then the number may still change, depending on what the typical frequency of the usage is. But if you are quite far away, then yes - they didn’t return.

View original

2 replies

Badge

For average unbounded retention , is this following a correct interpretation:

22.56% of users on average will return 29 days or later

But 13.23% of users on average will return 30 days or later

Thus 9.33% (22.56% minus 13.23%) of users on average return on the 29th day but not on the 30th day or later?

Badge

For your second question:

30-day unbounded retention means activity on 30th day or later.

1-month unbounded retention means activity in first month or later.

In the second case, everyone with activity in the entire first month is also counted. In the first case, the first 29 days won’t be counted.

For your second question:

The report does not show the prediction, but what is the current status. 9.33% of people who returned on 29th days, have not yet returned after that. If you are not too far away from that date, then the number may still change, depending on what the typical frequency of the usage is. But if you are quite far away, then yes - they didn’t return.

Reply