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

Optimize
2.4K members online now
2.4K members online now
Discover how to select who is eligible to be in your experiment, and when they’re served experiment variations.
Guide Me
star_border
Reply

Cancel a users test session on login.

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

I'm currently setting up an A/B experiment to track conversion of new customers. I determine wether a customer is new based on wether he/she is logged in or not.
However it might also be the case that an existing customer arrives on the checkout page, and then logs in. I don't want to include that customer in the test, how can I make sure he is excluded from the experiment? Does it happen automatically when the checkout page is reloaded and the javascript variable used to indicate to Google Optimize wether the customer is logged in or not, changes? Or can I do something programmatically to make sure to exclude the customer?

1 Expert replyverified_user

Cancel a users test session on login.

Google Employee
# 2
Google Employee

Hi

 

No I don't think that this will happen automatically, once a user has been assigned to a variant, he will be always considered part of that variant even if conditions later prohibit the variant from been applied again.

 

The main idea behind this is that once a user has seen the variant, he will be biased.

 

One solution to this problem, may be to define a new GA goal so that you consider conversions only those from non logged in customers.

 

In case you use GA events to track conversions and you have the logged-in property as part of the event data you may tryout the newly launched Optimize custom objectives (so that you don't have to create a new GA goal).

Cancel a users test session on login.

Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭

Thanks for the response. I can confirm that this does not happen automatically. I will look in to the solutions proposed and report my findings back here.

Thanks again!