Hi
We have published AB Test a month ago and the report shows version A more displayed than B with a significative gap (75 vs 25 on 100 visits).
How to explain this behavior since the activity is set with a 50/50 split. Can it be resolved automaticaly with an additionnal month running with same quantity of visits ?
Can it be forced to get more or less same number of impression at the end ?
Thx
Solved! Go to Solution.
Topics help categorize Community content and increase your ability to discover relevant content.
@Tienzen there could be multiple reasons for this behavior . May I know which reporting source are you working with ? If you are using A4T (Analytics as a reporting source) then this behavior could be because of unstitched hits . You may have a look at this document for more details around "unstitched hits/partial data" : https://experienceleague.adobe.com/docs/target/using/integrate/a4t/troubleshoot-a4t/minimizing-infla....
You may also look at the firing order of the experience cloud libraries on the page . The order should be Visitor ID > Target > Analytics .
If you are using Target as a reporting source , then please ensure you are not using any audience refinements that may lead to tis behavior.
For more detailed investigation you may create a ticket with the client care team using Admin console.
Hope this helps.
@Tienzen there could be multiple reasons for this behavior . May I know which reporting source are you working with ? If you are using A4T (Analytics as a reporting source) then this behavior could be because of unstitched hits . You may have a look at this document for more details around "unstitched hits/partial data" : https://experienceleague.adobe.com/docs/target/using/integrate/a4t/troubleshoot-a4t/minimizing-infla....
You may also look at the firing order of the experience cloud libraries on the page . The order should be Visitor ID > Target > Analytics .
If you are using Target as a reporting source , then please ensure you are not using any audience refinements that may lead to tis behavior.
For more detailed investigation you may create a ticket with the client care team using Admin console.
Hope this helps.
Views
Like
Replies