Analytics
2.2K 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

Experiment Sessions Too Low (VS regular Sessions)

[ Edited ]
Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

Hi there!

 

I have set up a site wide experiment according to this guide: https://developers.google.com/analytics/solutions/experiments-client-side#define

 

I have included 100% of my users, but the "Experiment Sessions" is less than half compared to the (regular) Sessions of the 4 past days. It's not even close to the number of Users. Check out: http://screencast.com/t/gxKjzFYWlTGm

 

I do have some Filters in place (exclude home and office IPs), but still: the Experiment Sessions should always be higher, right?

 

How are the "Experiment Sessions" defined, which can explain this difference?

Re: Experiment Sessions Too Low (VS regular Sessions)

Follower ✭ ✭ ☆
# 2
Follower ✭ ✭ ☆
Hi Erik,

If you don't mind, I'd like to have a look at the discrepancy. Please provide the following:
1) A URL to the site where the experiment is running.
2) The URL to the experiment report in GA.

Thanks, Jaren

Re: Experiment Sessions Too Low (VS regular Sessions)

[ Edited ]
Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭

Dear Jaren,

Thanks for your reply.

1) Check out https://networklessons.com  - It's the text before the button in the top bar -> http://screencast.com/t/i30NYc87tMD
2) To my knowledge, it is not possible to share an experiment report by URL in GA? If it is possible, please let me know how to do this :-)

Looking forward to your reply. Thanks!

Best regards
Erik

Marked as Best Answer.
Solution
Accepted by topic author Erik M
May 2016

Re: Experiment Sessions Too Low (VS regular Sessions)

Follower ✭ ✭ ☆
# 4
Follower ✭ ✭ ☆
Hi Erik,

1) I had a look at your site. It looks like you are loading the experiment API async:
jQuery.getScript("//www.google-analytics.com/cx/api.js ...

The variation needs to be chosen at least before your final GA tracking call. In my test visit to the site, the only GA tracking call occurred before the experiment API loaded, so the experiment was not included in the tracking beacon.

Have a look at this section for tips on synchronizing the API calls or what to do when you can't guarantee the order: https://developers.google.com/analytics/solutions/experiments-client-side#implement-send

2) It's not possible to share a report by URL. The report URL just happens to be helpful starting point for a GA representative to begin debugging. In any case, I believe that coordinating your experiment application prior to sending a GA hit should resolve your issue. You need not supply the report URL.

Good luck!
Jaren

Re: Experiment Sessions Too Low (VS regular Sessions)

Visitor ✭ ✭ ✭
# 5
Visitor ✭ ✭ ✭
Dear Jaren,

Thanks for taking a look and your help! :-)

Best regards
Erik

Experiment Sessions Too Low (VS regular Sessions)

Visitor ✭ ✭ ✭
# 6
Visitor ✭ ✭ ✭

Hi! Thank you for your solution. I have the same problem. Could you tell me please which debugger do I need to use in order to see this sequence?

I am using the same client side script within Google Tag Manager.

 

In GTM preview I see:

https://www.screencast.com/t/WrUFiZbiyv

 

In GA debugger I see:

https://www.screencast.com/t/038ly25o

 

I cannot figure it out weather JS code goes before or after ga tracking call.