Analytics
6.4K 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

Adjust Cohort Analysis (D14 and D30 retention)

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

Hi everyone,

 

is there a possibility to adjust the cohort analysis so that it does not only show D1-D12 Retention but also D14 and D30?

 

Best,

Jan

1 Expert replyverified_user

Re: Adjust Cohort Analysis (D14 and D30 retention)

Top Contributor
# 2
Top Contributor
Hi Jan,

I don't think that it's possible. Probably because these special reports in Analytics are constrained by the area allowed to fill in the web page due to usability and/or simplicity issues.

Christoph

Re: Adjust Cohort Analysis (D14 and D30 retention)

Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭
Hi Christoph,

thank you for your reply. Do you have any diea if there is a workaround? The usual mobile KPIs include (next to D1, D3 and D7): D14, D30, D60 and D90. If I cannot get these from GA, I will have to install another Analytics suite just for this.

Jan

Re: Adjust Cohort Analysis (D14 and D30 retention)

Top Contributor
# 4
Top Contributor
Hi Jan,

You might change cohort size to weeks and consider date ranges up to 12 weeks then.

Christoph

Re: Adjust Cohort Analysis (D14 and D30 retention)

Visitor ✭ ✭ ✭
# 5
Visitor ✭ ✭ ✭
Hi Christoph,

thank you for your reply! I experimented a little with weekly cohorts but I have a hard time understanding what it means exactly.

For example:
The cohort from 10.01.2016 - 16.01.2016 has 23 people and a week 1 retention of 8.7%. (8.7% translates into 2 people).
However, when I look at the daily retention rates of these cohort I can see that on day 1, one person came back. One day 2, a different person came back. And on day 6, one more person came back. This adds up to 3, not 2, people coming back in the first week.

Adjust Cohort Analysis (D14 and D30 retention)

Visitor ✭ ✭ ✭
# 6
Visitor ✭ ✭ ✭

Hey Jan,

 

Did you understand this mess? Smiley Happy

I'm encountering the same issue...

Adjust Cohort Analysis (D14 and D30 retention)

Visitor ✭ ✭ ✭
# 7
Visitor ✭ ✭ ✭

Hey Julien,

 

I could not find a satisfying answer by now. It may be a sampling error though. In the free version of GA data is often sampled and I think that a different sample is used for different time periods. Also I feel the sampling is not done very well, which leads to a large margin of error. So by now I decided that GA, in the free version, is relatively unsuitable to run A/B tests for large amounts of data.

 

Best

Jan