Analytics
1.7K members online now
Understand information in your reports and troubleshoot reporting issues such as self-referrals, (not set) data, and inaccurate information
 
Guide Me
star_border
Reply

Cohort analysis cumulative value calculation?

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

How are the cumulative values for Week 2 and Week 3 calculated? I understand that the Week 1 values are a weighted average, but it's unclear why the other values are so low when the underlying weekly values are much larger. 

 

As an example, here is what I see:

 UsersWeek 1 retentionWeek 2 retentionWeek 3 retention
Overall806749.0316.986
4/24/2016330445.9140.0132.23
5/1/2016241155.6637.54 
5/8/2016235246.6 

 

 

 

Screenshot link: https://www.evernote.com/l/AAN3TZ-YYbdCwYDf7Kz_wgrkvhoPb703oGsB/image.png

 

Any insight?

 

 

Re: Cohort analysis cumulative value calculation?

Visitor ✭ ✭ ✭
# 2
Visitor ✭ ✭ ✭
I have the same issue. The Overall numbers appear to be using a percentage of the overall total, instead of just the total for the weeks with data. This seems to me to be just plain wrong.

In your example, the "overall" number for week 3 should be 32.23%

Actually, I don't even know how they are calculating, because
(.3223 * 3304) / 8067 = 13%
so even the 6% value being displayed is wrong by that metric

Re: Cohort analysis cumulative value calculation?

Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭
I'm surprised no-one else has reported this issue; the calculations seem blatantly incorrect. Does anyone get value out of the cohort analysis in any other way?