Switch over from Manual Traffic Allocation to Auto-Allocation for the A/B Test | Community
Skip to main content
K
Karuppiah_Sakthivel
Level 4
November 1, 2023
Solved

Switch over from Manual Traffic Allocation to Auto-Allocation for the A/B Test

  • November 1, 2023
  • 1 reply
  • 1757 views

Hi Community,

 

Below is my scenario am trying to apply

 

I was run A/B testing for 18 days but it didn't announce the winner. So I have stopped. I would like to understand and apply Auto-Allocation. Below is the activity details and test result

 

Activity Details: Inserting new component about the product detail description.

Experience A: Control: 6600 Visitor, Conversion Rate: 14.74%

Experience B: 6500 visitor, Conversion Rate: 15.20%, Confidence Level: 53%

Average Lift: 3%

 

If I re-run the same activity by setting Auto-Allocation, Can i determines the winner more quickly . Suggestion please.

 

Thanks

KP

This post is no longer active and is closed to new replies. Need help? Start a new post to ask your question.
Best answer by Rajneesh_Gautam_

Hi @karuppiah_sakthivel  - definitely a candidate for auto-allocate.

 

Please note that the aim of auto-allocate is not to find the winner quickly - its more to drive traffic to winning experience so that business may get most of the experiment, rather than waiting for the experiment to finish.

 

I'd recommend switching to auto-allocate if your objective is to get more benefits (sign-ups, sales, revenue....) even while the experiment is running. Typically this should be done when you experiment life is not very long e.g. testing different messages on a marketing campaign. 

 

Hope this helps

 

Regards

Rajneesh

1 reply

Rajneesh_Gautam_
Community Advisor
Rajneesh_Gautam_Community AdvisorAccepted solution
Community Advisor
November 1, 2023

Hi @karuppiah_sakthivel  - definitely a candidate for auto-allocate.

 

Please note that the aim of auto-allocate is not to find the winner quickly - its more to drive traffic to winning experience so that business may get most of the experiment, rather than waiting for the experiment to finish.

 

I'd recommend switching to auto-allocate if your objective is to get more benefits (sign-ups, sales, revenue....) even while the experiment is running. Typically this should be done when you experiment life is not very long e.g. testing different messages on a marketing campaign. 

 

Hope this helps

 

Regards

Rajneesh

    K
    Karuppiah_SakthivelAuthor
    Level 4
    November 1, 2023

    Hi Rajneesh,

     

    Thank you for the suggestion, Can i re-run the same activity?

     

    How It will work? Based on the previous confidence level on the existing report.

     

    Thanks

    Karuppiah

    Rajneesh_Gautam_
    Community Advisor
    Rajneesh_Gautam_Community Advisor
    Community Advisor
    November 2, 2023

    Hi @karuppiah_sakthivel  - if you are starting a new activity, everything will be start from scratch which means results of your previous activity will not have any impact.

     

    Regards

    Rajneesh

    Terms & ConditionsAccessibility statement

    Loading footer...