AdWords is now Google Ads. Our new name reflects the full range of advertising options we offer across Search, Display, YouTube, and more. Learn more

Analytics
4.3K members online now
4.3K members online now
Dive into multiple domain (Cross/Sub) tracking, implementing Ecommerce and Enhanced Ecommerce, setting up Event tracking, and Universal Analytics code.
 
Guide Me
star_border
Reply

experiment

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

I have an experiment running and have set the traffic to 100%. The numbers I'm actually getting are around 17%.

 

Here's what the experiment says:

The experiment, homepage_content, will evaluate Homepage Click Through (Goal 5 Completions) on 100% of user traffic.
Email notification is set for this experiment.
2 weeks minimum run time, 99.5% confidence threshold.
 
I'm scratching my head and would love to know what to know what I"ve done wrong?
 
Thanks in advance
1 Expert replyverified_user

experiment

Rising Star
# 2
Rising Star

Hi James,

 

How long has it been running?  Also for anyone else, 100% means 50% will see the original and 50% will see the alternate... at least at first.

 

Best,

 

Theo Bennett

Analytics Evangelist at MoreVisibility | Contact Me
Connect on LinkedIn

experiment

Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭

Hi Theo

It's been running for 10 days. There are 3 variants. The total sum of all traffic to the experiment is about 17%. So what I'm seeing is roughly 33% of the  17% for each variant, rather than the 100% split across the 3

 

Thanks

James

 

experiment

Rising Star
# 4
Rising Star

Hi James,

 

You should have (on the first day) 25% to each page (if you have an original or "A" page and 3 variants or 33% if the A page and 2 variants). 

 

GA will then re-balance the percentages towards one that is more likely to convert and this could be why you see the percentages waiver. 

 

https://support.google.com/analytics/answer/2844870?hl=en

 

If you don't have a lot of traffic, 3 variants could take awhile.

 

Best,

 

Theo Bennett

Analytics Evangelist at MoreVisibility | Contact Me
Connect on LinkedIn

experiment

Visitor ✭ ✭ ✭
# 5
Visitor ✭ ✭ ✭

Hi Theo

Thanks for the response.

 

For clarity, there are 3 variants in total.The current home page is one of the options.

We set the experiment to run with an even distribution, overriding the default setting of pushing more traffic to the one showing the most promising results.

The traffic to the experiment is set to 100%.

 

The site received 1027  users on Thursday.  

The experiment  56, 62, 50  users to each variant. 

 

I'm thinking of stopping the experiment. Copying it and starting from fresh. But any ideas are welcome.

 

Thanks

James

experiment

Visitor ✭ ✭ ✭
# 6
Visitor ✭ ✭ ✭

Hi Theo

figured it out. 

The traffic allocation is working properly. What wasn't working properly was my knowledge of visits that include the homepage. This number is much lower than I'd assumed. So I'm a little embarrassed, but thanks for your help.

 

James 

experiment

Rising Star
# 7
Rising Star

Hey James,

 

No worries!  Smiley Happy

 

-Theo

Analytics Evangelist at MoreVisibility | Contact Me
Connect on LinkedIn