Solved

Average Time Between Sessions

  • 16 September 2021
  • 4 replies
  • 656 views

Badge

Hi everyone !
We’re trying to find the average time between sessions ie After how long does a user come back to the platform.
I know this sounds just like retention but we don’t want that because from what I understood , and correct me if I’m wrong, the retention analysis tells that for all users who did event A on Day 0 , how many did event B on Day N .

We’re just looking for time difference from when people log out (end session) and then log in (start session) and if possible per day.

Would you recommend any of the following options : 
1. Seeing the Time to Convert in a Funnel with first event as End Session & second event as Start Session
2. N-Bound Retention analysis with the same event order

Hope to hear back from the community !

icon

Best answer by jmagg 17 September 2021, 17:25

View original

4 replies

Badge

Curious to know this one.

 

Userlevel 5
Badge +6

Hi @apo1397 - I’d say option 1 is the best way to calculate the time difference from when people log out (end session) and then log in (start session). Here is a chart I’ve made in the demo environment to illustrate. Feel free to adjust the “..completed within” metric to days, hours, minutes, or seconds & change the date picker to isolate the days of interest. Then, like you mentioned, scroll down to the breakdown table to see the average & median times it took for users to go from End Session to Start Session. Do you think this will work in your case?

Badge

Thanks for the reply @jmagg ,
I went ahead and changed the metric from conversion to Time to Convert here
So according to this 18% people logged back in within 10 mins ?
Is there a way to get this as a number (avg or median) only ?

Userlevel 5
Badge +6

Hi @apo1397 - no problem! At the moment, there is not a way to extract the Time to Convert values to avg/median numbers, but you can play around with the bin & interval sizes to zero in on specific time ranges. For example, see this chart here looking at just those users who fall between 1m & 10m. 

Reply