FHR - A/B Tests - How to: create a result modification A/B test.

Function

This document will explain the process for creating an A/B test using result modifications.

Scenario

Steps 

1. Setup an A/B Test using Result Modifications 
   
   •  Login to the Merchandising studio.
     
     
   • Select A/B test Menu.
     
     
     
   • From the A/B test screen, Click New to create a new test
     
     
     
   • The New test screen appears
     
     
     
   • Select the A/B test type using the drop-down as 'Result Modification' 
     
     Once selected, click the select scope button
     
     
   • From the scope screen, select the relevant scope for the test. (here we will be using 'en_GB')
     
     Then click the select button to confirm.
     
     
   • Now select the trigger type to be used 
     
     
     
   • Select the trigger for the A/B test. (For this example 'user location' will be used)
     (add screenshot with only the standard triggers showing)
       
   • Now the trigger is selected, click 'add location' button 
     
     
     
   • The modification screen opens to allow the location to be set
     
     
     
   • In this example, we will use the 'clothing' category by selecting it from the category facet. 
     
     
     
   • Once the location has been set, click the 'select location' button to return to the previous screen.
     
     
     
   • The 'User location' can be adjusted if required for contains, equals, ends_with etc.., 
     
     Scroll down to the first case 'Case A' to continue setting up the test.
     
     
   • Select if case A will use the underlying page ranking or not.
     
     
     
   • In this example, we will select 'use modifications' from the drop-down.
     
     
     
   • Click 'Configure modifications'.
     
     
     
   • This will open the modifications screen.
     
     
     
   • Click to create a new group,  (This example will group buying_season 'SS22')
     
     
     
   • Select "Create group":
     
     
     
   • You can now add 'Rank by' filters if required and select the number of items for the modification. (In this example we will set it for 10 items to be included).
     
     
     
   • Click the 'Apply' button.
     
     
     
   • A new window opens, and the group is now shown and highlighted.
     
     
     
   • Select the icon to see the group settings.
     
     
     
   • Select what you wish to do with the group (In this example we will move the group to the bottom of the page). 
     
     
   • The screen will refresh. Select >> to be taken to the bottom of the page.
     
     
     
   • Here we can see the group has been moved to the bottom, click back one page to see the remainder of the group.
     
     
     
     
   • Once happy with the position, Click Finish to confirm.
     
     
     
   • You will be returned to the A/B test screen.
     
     You can now add more modifications to 'Case A' if required (we will not for this example).
     Also, the traffic allocation split can be adjusted here (50% is the default) we will leave it as is.
     
     
   • We can now edit the settings for 'Case B' 
     
   • Click the 'Configure modifications' button 'Case B' to open the modifications screen.
   • The location 'clothing' will already be set based on the A/B test trigger we set earlier. 
     
     
   • Click on the group,  (For 'Case B' we will also group 'buying_season 'SS22' )
     
     
     
   • Select to create the group
     
     
     
   • You can now add 'Rank by' filters if required.
     
     
     
   • Set the number of items for the modification to 10 items.
     
   • Click the 'Apply' button.
     
     
   • The result modification screen opens and the group is now shown and highlighted.
     
     
     
   • Select the icon to see the group settings.
     
   • For 'Case B' we will move the group to position 1 by pressing the 'Apply' button.
     
     
   • The group's position will remain the same as it is already at the top of the 'clothing' section, Click Finish to confirm.
     
     
     
   • You will be returned to the A/B test setup page.
     
   • Further test cases can then be added to make the test an A/B/C test. (This example will stick with A/B).
     
   • Once happy with the test cases, click to save the test.
     
     
     
   • Now you can 'send for publication'
     
     
   • You can now see the test listed on the A/B test page.